66 67

Variables Characteristics

Location Hofstetten, 12 km south of the city of Basel in NW Switzerland47°28' N, 7°30’E

Altitude 550mMean annual air temperature 10°CMean annual rainfall 990mmType of forest Mixed coniferous temperate forestArea of forest accessed by the crane 0.28ha (note: the crane is located in a very small natural gap)Canopy height 32-38mCrane model Liebherr 30LC, fixedHeight of tower / Length of jib 45m/30mMaximum height reached by the gondola 37mGondola type a: Cylindrical, model RM1-300A/32; 65cm in diameter

b: Square, model SEC 02/600; 1.2 x 1.2mNumber of persons carried by the gondola a: 1 person

b: 4 personsIn operation since 1999Main research topics • Tree responses to CO2-enrichment in the canopy

• Responses of functional leaf types to light and shade in relation to forest canopy structure

Remarks Fenced research site, equipped with laboratory, power and telephone linesManagement Institute of Botany, University of BaselContacts Prof. Christian Körner, University of Basel, Ch.Koerner@unibas.ch

Dr Gerhard Zotz, University of Basel, gerhard.zotz@unibas.chWeb site www.unibas.ch/botschoen/scc/List of publications www.unibas.ch/botschoen/scc/Fees for researchers Negotiable on a case to case basis

Background

The Swiss Canopy Crane (SCC, Fig. 19 and Table 2) was erected with

the help of a helicopter in March 1999 at Hofstetten, close to the town

of Basel. It is managed by the Institute of Botany, University of Basel,

and is sponsored by the Swiss Federal Office of the Environment

(BUWAL), the Swiss National Science Foundation (SNF), and the

University of Basel. On the fenced research site (circa 1ha), there is a

field laboratory, power and telephone line. In September 2000 a large-

scale canopy enrichment system (web-FACE: Free Air CO2 Enrichment)

was installed, which permitted exposure of 14 adult trees to elevated

levels of carbon dioxide (CO2; a detailed description of this system can

be found in Pepin & Körner, 2002).

4.2. Cranes in temperate forests4.2.1. Basel, SwitzerlandChristian Körner and Gerhard Zotz

Fig. 19. View from the forest floor to the top of the crane, which was set upin a small natural gap of a few m2.

Table 2. Site and crane characteristics of the Swiss Canopy Crane.

The SCC is administered by the Institute of Botany, University of Basel. Three technicians ensure day-to-

day operation of the crane, the web-FACE system and the monitoring of macroclimatic and microclimatic

variables. The crane site was primarily selected to allow comparative studies of the impact of elevated

CO2 on mature individuals of typical European forest tree species. In addition to questions related to

global change, the high tree species diversity at the site provides a unique opportunity to study various

other topics in plant sciences, entomology, or forest pathology (Hoch et al., 2003).

Japanese Government. Eventually, in 2001, two new cranes were erected in temperate forests of Germany, in

Leipzig and Freising, respectively.

The network now consists of six cranes erected in temperate forests and five in tropical forests (Fig. 18).

They will be shortly joined by the Canopy Operation Permanent Access System (COPAS) in French Guiana,

a fixed device with a different conception (Chapter 4.3.2). These sites are located in forests from different

types, biomes and biogeographical regions, such as northern coniferous forest, mixed temperate forest,

decidous broad leaf forest, tropical dry lowland forest, and tropical wet lowland forest. To date, no crane site

has been established in Africa.

The International Canopy Crane Network was founded in 1997 (Stork et al., 1997a), during the organisation

of a Tropical Forest Canopy Symposium held in Panama in March 1997, and in response to an earlier call to

promote the long-term studies of forest canopies (Parker et al., 1993). Sixty-one participants representing 23

nations attended the meeting in Panama, including delegates from UNEP, UNESCO, CIFOR, and IUCN.

This meeting complemented a series of International Canopy Conferences held in Sarasota, USA, in 1994

and 1998; and in Cairns, Australia, in 2002, with the next meeting planned for Leipzig, Germany, in 2005.

Each crane site has unique peculiarities and its associates are involved in different research topics. Yet some

baseline investigations are common to all sites: identifications of plant and animals present, mapping and

measurement of trees, microclimatic studies, etc. The ‘Global Canopy Handbook’, edited by the Global

Canopy Programme (Mitchell et al., 2002), presented the different crane sites of the network, particularly in

terms of technical characteristics and costs of installations. In the following sections, the manager(s) of each

crane research facility were asked to describe the climate and vegetation of their site and to provide an

overview of the past, present and future research at their facility. These findings are further summarized in

Chapter 5 and put into the perspective of global canopy research.

68 69

The SCC is located in a typical low-altitude mixed forest of central

Europe. The forest is about 100 years old, with tree heights ranging from

32 to 38 m, and a total basal area of 46 m2 ha-1 (Table 3). The stand is

characterized by a dominance of beech (Fagus sylvatica) and oak (Quercus

petraea/robur) with representatives of lime (Tilia platyphyllos), hornbeam

(Carpinus betulus), maple (Acer campestre), and wild cherry (Prunus

avium). There are also four coniferous species, European larch (Larix

decidua), Norway spruce (Picea abies), Scots pine (Pinus sylvestris), and

silver fir (Abies alba). Only the latter gymnosperm is currently

regenerating while the other three species became established in the late

19th century (they are more light-requiring species). The spatial

distribution of tree species within the crane’s perimeter is shown in Figure

20. Finally, the site has a strong presence of ivy (Hedera helix) reaching

the upper canopy. The leaf area index (LAI) of the forest in the crane

plot is circa 5.

There is also a rich understorey shrub flora with hazel (Corylus avellana),

honeysuckle (Lonicera xylosteum), spurge laurel (Daphne laureola), and the evergreen holly (Ilex

aquifolium). The herbaceous layer is dominated by Mercurialis perennis. Other important species include

Paris quadrifolia, Anemone nemorosa, and Galium odoratum. Additional tree species are found in the

immediate vicinity of the crane perimeter, for example, ash (Fraxinus excelsior), whitebeam (Sorbus aria),

and a second species of maple (Acer pseudoplatanus). Thus, a total of 16 woody species reach the upper

forest canopy.

The regional climate is typical for the humid

temperate zone, characterized by rather mild winters

and moderately warm summers (Fig. 21). The

growing season of the deciduous trees lasts circa 170-

180 days from the end of late April to mid-October.

Average minimum January and maximum July air

temperatures near the crane site are -1.4 and 23.9°C,

respectively. Annual precipitation in the region

averages 990mm, two-thirds of which falls during the

growing season. Soils are of the rendzina type on

calcareous bedrock (a silty loam with an accessible

profile depth of circa 30cm and a pH of circa 6 in the

top 10cm of the profile).

Main research topics and first findings

The central objective of the SCC is the investigation

of the responses of mature trees to elevated CO2, and

consequently, the web-FACE system constitutes the

‘heart’ of the facility. Briefly, pure CO2 is released

through thin tubes, which are woven into the canopy.

The set point CO2 concentration is controlled by a pulse-width

modulation routine. About 2 tonnes of CO2 per day are needed

for 14 tall forest trees (Fig. 20, thick circles). By now we have

completed two full growing seasons, in which these trees have

been exposed to an average of 510ppm CO2, which mimics

ambient CO2 concentrations of the near future. During this

exposure to high CO2 we assessed the response of six different

tree species by studying leaf gas exchange parameters (Fig.

22), leaf chemical composition, stem sap flux, stable isotopes,

tree phenology, increments in diameter at breast height and

branch growth patterns.

The first results indicate a number of significant responses to

elevated CO2 in spite of considerable intraspecific and

interspecific variation. For example, in broad-leaved trees leaf

stomatal conductance was reduced by circa 15% on average (Fig. 23); range: -3% in Fagus sylvatica to -

22% in Carpinus betulus) compared to controls at normal CO2 levels, although differences were only

significant in two species. None of the conifers showed a significant difference. Consistent with these

results, sap flux data for the same year suggest a reduction of transpiration by almost 10%. Conversely,

tree growth showed no clear trend up to the present but much variation between species and years.

Cooperating teams from a number of institutions have examined, for example, isoprene emissions, insect

abundances, or the responses of herbivores to leaves that had developed under elevated CO2. For example,

preliminary experiments on the feeding behavior of the generalist herbivore, Lymantria dispar, yielded

Stems Basal areaTree species per ha (m2 ha-1)

Abies alba 57 1.26Acer campestre 18 1.22Acer pseudoplatanus 35 0.11Carpinus betulus 64 2.94Corylus avellana 11 0.02Crataegus sp. 3 0.01Fagus sylvatica 106 11.02Larix decidua 42 8.71Picea abies 67 5.82Pinus silvestris 21 3.26Prunus avium 7 1.13Quercus petraea/robur 67 8.46Tilia platyphyllos 117 2.37TOTAL 615 46.32

Table 3. Stand characteristics by tree species(measured 1999).

Fig. 20. Tree positions at the SCC site.

Fig. 21. Climate diagram for the SCC site Metzerlen, near Basel/Hofstetten. Averages are based on 1989-1999meteorological records.

Fig. 22. Sonja Keel, a master studentmeasuring leaf stomatal conductance oncanopy leaves. Standing: Olivier Bignucolo.

70 71

Background

Canopy research is being conducted in a managed forest of beech and spruce (Fagus sylvatica and Picea

abies) in southern Germany, about 40km north of Munich. The study site, Kranzberg forest, includes

about 500 spruce and beech trees that form a closed canopy (Pretzsch et al., 1998). Trees are about 60

years old, and the foliage extends from 17m aboveground to the upper canopy edge at 28m (Table 4).

Besides the two dominating tree species, individuals of pine (Pinus sylvestris), larch (Larix decidua), oak

(Quercus robur), and maple (Acer pseudoplatanus) are scattered in the forest plot. Access to the canopy is

provided by a research crane and scaffolding (Fig. 25):

- The stationary research crane (KROCO-Kranzberg ozone canopy observation by crane) was installed

in April 2001, is operated by permanent power supply and permits sampling and measuring across

the entire study site (Table 4).

- The scaffolding consists of four towers, ranging between 27 and 35m in height. Three towers are

connected by platforms (12m in length) that allow access at four heights between 17 and 25m above

ground to the insolated and shaded crowns of about 30 trees (installed in December 1996). The

construction enabled the installation of a unique free-air fumigation system (KROFEX-Kranzberg

ozone fumigation experiment) to study the response of mature forest trees to long-term ozone

exposure at an enhanced level (Werner & Fabian, 2002; Häberle et al., 1999), an experiment which

started in 2000.

Further, equipment for the generation and control of ozone and for the

assessment of environmental factors, an ecophysiological field laboratory

(analysis of leaf gas exchange, sap flow in xylem, stem, branch, root and

soil respiration, growth rates and biomass production), and a central

laboratory hut belong to the infrastructure of the experimental site.

The research in the Kranzberg forest is maintained by around 25 national

and international groups covering the fields of molecular biology,

biochemistry, ecophysiology, plant nutrit ion, mycorrhizae,

phytopathology, zoology, soil science, forestry, air chemistry and

modelling, half of them from outside the Munich area. There is interest

in extending cooperation with external groups that may contribute in a

complementary way to the investigations conducted at the ‘Kranzberg

Forest’. A core group is meeting every month to coordinate the activities.

The main projects running at the site are:

(1) “Growth and Parasite Defense - Competition for Resources in Economic Plants from Agronomy and

Forestry” (s ince 1998, funded by ‘Deutsche Forschungsgemeinschaft ’ (DFG),

Sonderforschungsbereich ‘SFB 607’, www.sfb607.de);

(2) “Risk Assessment of the Enhanced Chronic O3 Exposure by Means of ‘Free-Air’ Canopy Fumigation in

quite contrasting responses to altered food quality (Fig. 24). This result

highlights once more the importance of a multi-species approach to study

biotic reactions to global change. Yet another methodological focus was

stable isotope work. Changes in δ13C signals under elevated CO2 (due to

the fossil CO2 source) could be traced from the canopy down to the

mycorrhizal fungi.

The knowledge of short-term reactions to elevated CO2 may have little

bearing to an understanding of long-term responses (Körner, 1995). Thus,

it is essential that the web-FACE system will operate for at least another

five years. Only a prolonged duration of our experiment will permit the

distinction of transient reactions from ecologically relevant long-term

responses of these mature forest trees. High biodiversity being one of

the assets of the SCC site it is also a burden in terms of adequate

replication. Thus, it is highly desirable to install replications of the

pioneering web-FACE system in a number of other locations within the

temperate zone, but also in other forest biomes around the world

considering the very nature of global change. Forest ecosystems contain

more than 80% of biomass carbon stored in the terrestrial biosphere.

An understanding of the response of these systems to global change is a

major challenge to modern biology.

Fig. 23. Differences in leaf conductance to water under ambient and elevatedCO2. Data are means ± SE.

Fig. 24. Differences in growth rate ofLymantria larvae feeding on leavesgrowing under ambient (A) and elevated(E) CO2. Data are means ± SE.

4.2.2. KROCO, Freising, Germany:Canopy research in a temperate mixed forest of Southern GermanyKarl-Heinz Häberle, Ilja M. Reiter, Angela J. Nunn, Axel Gruppe, Ulrich Simon,

Martin Gossner, Herbert Werner, Michael Leuchner, Christian Heerdt, Peter Fabian &

Rainer Matyssek

Fig. 25. Installations at Kranzberg Forest, D: Canopy crane, towers andplatforms with the free-air ozone fumigation system.

72 73

• Determination of eco-physiologically meaningful threshold levels in the O3 sensitivity of adult forest

trees and mitigation of their carbon sink strength under enhanced, chronic O3 exposure.

In parallel to the intensive investigations in the tree crowns the belowground ‘canopy’ of the roots has

recently become a new focus in the research at the site.

The utility of the canopy crane is exemplified by the following three multidisciplinary research activities.

Ecophysiology: light as driving force of competition

Our approach to quantify competitiveness is based on resource flux. Competitiveness is assessed as a

hierarchy of efficiency ratios in space sequestration (resource investment per unit of occupied above and

belowground space), resource acquisition (resource gain per investment and occupied space), and ‘running

costs’ (transpiration, respiration per resource gain and occupied space). In this way, costs versus benefit

relationships are established in the control of resource allocation within and amongst trees as being in

relation to exposure to biotic and abiotic impacts (Grams et al., 2002). Here, ozone is used as a perturbant

rather than pollutant to unravel the responsiveness and differentiation of such relationships as well as

their underlying mechanisms. Another fundamental factor being studied

is the naturally occurring light gradient across the canopy and its

influence on such costs versus benefit relationships. Additionally, the

impact of pathogens, mycorrhizospheric organisms and phytophagous

insects is of paramount importance with respect to the rationale of the

interdisciplinary research program ‘SFB 607’ (see above and Matyssek

et al., 2002). This implies that impact levels of parasites and their effects

on the trees’ primary and secondary metabolism be assessed, and that

the abundance of insects is surveyed in the stand’s canopy.

Tree species can differ distinctly in crown structure and in their strategy

to exploit canopy volume (Küppers, 1994). This is shown for leaf area

densities (LAD) in the vertical profiles of Norway spruce and European

beech (Fig. 26). Beech had a pronounced maximum of its LAD at the

top of the crown. LAD of spruce was more evenly distributed and

increased towards the crown base up to a maximum, which was one third

lower in comparison with beech. The irregularities in LAD of spruce

reflected the crown structure, where the growth pattern of branches

leaves gaps, that allow more light than in beech. At the stand level, beech

only had one third of the one sided leaf area index (5.6m2 m-2) and only

half of the projected leaf area index (10m2 m-2) in comparison with spruce

(15m2 m-2).

Morphological differences between sun and shade branches as well as biotic and abiotic injury to the

foliage are dependent on light exposure (Fig. 27). Before autumnal senescence the percentage of damaged

and shed leaf area increases in the insolated crown of beech more rapidly in comparison with that shaded.

In particular, necrotic areas on leaves in the insolated crown reduce the photoynthetic carbon gain

substantially, starting at the end of August. Cicadellidae (leafhoppers) are feeding preferentially on the

epidermal layer of shade leaves. Combining datasets from biochemical, ecophysiological and zoological

surveys will contribute to a mechanistic understanding of the observations.

a Mixed Beech/Spruce Forest” (since 2000, funded by ‘Bayerisches Staatsministerium für

Landesentwicklung und Umweltfragen’);

(3) “The Carbon Sink Strength of Beech in a Changing Environment: Experimental Risk Assessment of

Mitigation by Chronic Ozone Impact (CASIROZ)” (since 2002, funded by the European Community

within Fifth RTD Framework Programme EVK2-2002-00165 (Ecosystem Vulnerability),

www.casiroz.de).

The experimental site belongs to a forest that is ‘public property’ (i.e., owned by the Federal State of

Bavaria) and provided free of charge by the ‘Bayerisches Staatsministerium für Landwirtschaft und

Forsten’.

In an interdisciplinary approach from the gene to the stand level, the project intends to clarify central

mechanisms in plants, regarding the regulatory control of competitiveness and individual plant fitness.

In particular, the following research aims are pursued:

• Quantification of competitive interactions between adult beech and spruce trees as based on the

resource fluxes involved;

• Clarification of the associated resource allocation, employing experimental ‘free-air’ ozone (O3)

exposure within the canopy to facilitate the analysis and investigation of regulatory mechanisms and

their responsiveness;

Variable CharacteristicsLocation Kranzberg Forest, Freising, Bavaria, Germany

48º25’08" N, 11º39’41" EAltitude 485mMean annual air temperature 7.0 - 7.5°CMean annual rainfall 730 - 790mmType of forest Managed mixed spruce/beech forestArea of forest accessed by the crane 0.8haCanopy height 28mCrane model Potain, fixed, MDT-1sHeight of tower / Length of jib 45m/50mMaximum height reached by the gondola 35mGondola type Cylindrical, 0.7m in diameter, operated by remote controlNumber of persons carried by the gondola 2In operation since 2001Main research topics • Quantification of competitive interactions between adult beech and spruce

trees• Study of the regulation of carbon allocation using ozone exposure as an

experimental tool of disturbance• Determination of eco-physiological threshold levels of ozone sensitivity of

mature forest trees• Temporal and spatial contribution of spruce and beech to the faunal

biodiversity in the canopyRemarks Fenced research plot of 0.5ha, including the crane and four scaffolding

towers (27 to 35m tall) connected by platformsManagement Technical University of MunichContact Prof Dr Rainer Matyssek, Technische Universität of München,

matyssek@wzw.tum.deWeb site www.sfb607.deList of publications www.sfb607.deFees for researchers On demand

Table 4. Site and crane characteristics of KROCO in Freising.

Fig. 26. Leaf area densities (LAD) in the vertical profiles of Norway spruce(Picea abies L. [Karst.], n=3) and European beech (Fagus sylvatica L., n=7).Measurements were registered electronically with an optical instrument (LAI-2000, Li-cor Ltd., Lincoln, Nebraska, USA). The measurements on spruce wereconverted into projected needle area by multiplication by 3.59*(π/2)-1 (Fassnachtet al., 1994). The one-sided needle area (Chen & Black, 1992) was calculatedwith respect to needle morphology, which depended mainly on needle ageand position in the crown.

74 75

of 14m (shaded crowns), 20m (insolated crowns) and 26m aboveground (directly above the canopy)

demonstrate that the entire fumigated crown zone follows approximately ‘2x ambient ozone’. Vertically,

the ozone distribution within the canopies is homogeneous within about 20%. No fumigation is performed

below the stand canopy. Thus, the control instrument mounted at 6m aboveground does not display

elevated ozone levels.

A large number of passive samplers (Werner, 1992; Baumgarten et al., 2000), up to 200 being employed

simultaneously, were exposed within the fumigation zone as well as outside, in order to monitor, with

high spatial resolution, the ozone distribution in the stand. We used a sampling time of 7 days and, thus,

obtained 1-week integrations. As a typical example of the results obtained, the integrated ozone exposure

at 20m above ground (insolated crown level) is shown for one week in summer (Fig. 31). Isolines are

shown of the extinction of isatine, the product of ozone reaction with the Indigo-dye of the passive

Bioclimatology: the challenge of a ‘free-air’ ozone fumigation system

A novel system for continuous and controlled free-air fumigation of

mature tree canopies with ozone was set in operation in May 2000 (Figs

28 and 29). Within a volume of 2000m3 which comprises the crowns of

10 neighbouring trees, the O3 levels that prevail at the forest site are

experimentally increased to a ‘2 x ambient’ O3 regime (up to maximum

levels of 150ppb O3; Werner & Fabian, 2002). Five trees each of spruce

and beech are exposed to this regime, whilst another group of five of

each grow under the ambient O3 regime (i.e., unchanged O3 levels) to

serve as a ‘control’.

Ozone is produced by a commercial ozone generator (Ozonia-CSI), the

ozone delivery of which can be regulated between 0 and 70 g/hour. To

prevent formation of oxides of nitrogen, the ozone generator is operated

with oxygen rather than air. With a commercial oxygen generator based

on the pressure swing absorption (PSA) technique via a molecular sieve,

oxygen is enriched to 90% in air (including passage through a dryer and

VOC filter). The use of a 6 to 8 bar compressor and an air flow of about

470l/min proved to be an efficient and low-cost solution. The ozone

generator output is fed into a 2000l mixing tank, with a constant flow

rate of 1500l/min of ambient air, being added by means of a blower that

maintains a tank pressure of 1.2 bar. Differing from the FACE design

(Free Air CO2 Enrichment, see Karnosky et al., 2001, Pinter et al., 2000),

which utilises a ring-shaped tube system for fumigation that encircles

the experimental area, we use a system of 130 PTFE tubes fitted into

the mixing tank by a manifold to conduct the ozone/air mixture directly

into the canopies of the study trees. These tubes are fixed in a grid

mounted above the canopies, and hang downward, about 80 to 100cm

apart from each other. Each tube is equipped with 45 flow-calibrated

outlets, 33cm apart from each other, each providing a constant flow rate

of about 0,30l/min each.

This methodology ensures an experimentally enhanced and chronic

whole-tree exposure to ozone whilst avoiding, in the absence of plant

enclosure in chambers or cuvettes, physiological bias through micro-

climatic artefacts (which prevail in conventional fumigation studies). The

striking advantage is the applicability to adult trees growing in naturally

structured forest stands (Karnosky et al., 2001). The technique is suitable

for CO2 fumigation as well.

Ozone records (1-hour averages) at 4 heights within the non-fumigated control area are shown in Figure

30 for two weeks in August and September 2000, respectively, along with the reference background

ozone level as recorded above the stand canopy at about 50m distant. This background level is not

influenced by the fumigation experiment, as indicated by the high correlation with the ozone level recorded

at the roof of our institute building, 5km distant from Kranzberg forest. The 1-hour averages at 3 levels

Fig. 27. Percentage of damaged leaf area in the insolated and shaded crownof European beech.

Fig. 28. View on the ‘free air’ ozone fumigation system(KROFEX).

Fig. 29. Schematic diagram showing the principle components of the KROFEX design.

76 77

Fig. 31. Distribution of ozone doses cumulated during one week (26.vii.-01.viii. 2000) by passive sampling in the sun crowns 20m above ground level.The fumigated trees (‘2 x O

3’) are marked by italic, the control trees (‘1 x O

3’)

by bold letters: B = Beech, S = Spruce. Positions of samplers are shown bysmall circles, isolines of extinction were derived by Kriging’s method.

Fig. 32. Branch beating to collect arthropods in winter.

Fig. 33. Means of spider abundance per tree (beating of two branches pertree, see text) in samples from beech (n=10) and spruce (n=10) during winter.Note logarithmic y-axis.

samplers. This extinction is proportional to the ozone deposition during the exposure time of the samplers

and therefore a measure of the integrated external ozone dose of the particular week.

Extinction values of 0.45 to 0.50 prevail within the fumigation zone (Beech = B, spruce = S, Fig. 31).

Non-fumigated reference trees (bold letters B and S) are shown, most with extinction values around

0.25. Thus, as is noticeable from Figure 30 as well, ambient ozone in this example was increased by about

a factor of 1.8, well within the majority of the other data (not shown), ranging between 1.7 and 2.0.

The concept focuses on the functional performance of tree individuals growing in stands with their multi-

factorial biotic and abiotic interactions. The comparison between the two O3 regimes enables a quantitative

risk assessment of a broad spectrum of molecular, biochemical and ecophysiological tree responses under

the given site conditions. This allows the examination of tree processes which are at risk or already

reflect incipient injury under the unchanged O3 regime. Each of the study trees is viewed, statistically, as

an individual case study, with the intention of deriving consistency patterns from O3 responses occurring

synchronously at the cell, organ and whole-tree level. Mechanistic modelling is used in scaling findings

to the stand level. Findings aid the definition of missing, field-relevant measures of O3 sensitivity in adult

forest trees and will make thresholds based on O3 exposure obsolete while promoting concepts of actual

O3 uptake, i.e. the O3 flux into leaves through stomata. By this, physiologically and ecologically meaningful

O3 doses are provided with respect to tree acclimatisation to ozone, mitigation in carbon sink strength of

trees and stands, and initiation of O3 injury (Häberle et al., 1999).

Summarizing recent results from the on-going experiment, beech leaves

developed visible symptoms and accelerated autumnal senescence due

to the elevated ozone regime whereas spruce appeared to be less

susceptible (Nunn et al., 2002). An important finding is that responses

from containerised young plants cannot be extrapolated unconditionally

to the performance of adult trees under ozone stress.

Animal ecology: the spatial and temporal distribution of biodiversity intree crowns

The primary aim of our studies is to assess the role of coniferous and

deciduous trees as habitats for arthropods in a managed forest, and to

examine the role of these tree species for the maintenance of biodiversity

in managed forests. Bavarian experience in arthropod ecology of tree

crowns derives from the comparison of communities of managed and

unmanaged forests at different sites (Simon, 1995, 2001, 2002; Schubert,

1998; Ammer & Schubert, 1999; Gruppe & Schubert, 2001; Gossner &

Simon, 2002). The methods have been adapted successfully to the

conditions of the Kranzberg forest, facilitated by the canopy crane. Thus,

the studies of the arthropod fauna could be extended to the upper and

outer canopy of spruce and beech, including seasonal replication. In order

to work non-destructively we decided to sample arthropods by branch

beating (Fig. 32). We use a funnel with an area of 0.5m2, sampling two

branches (each more or less covering the funnel opening) of one tree in

the upper and outer canopy by standardised beating (ten beats per

branch). Sampling was performed in ten spruce trees and ten beech trees

in monthly or, during winter, bimonthly intervals.

A bottleneck for the survival of arboreal arthropods in temperate forests

is apparently winter time. In exposed tree crowns climatic conditions

are even worse than on ground level where the forest cover mitigates

extremes in temperature. Our results indicate that all arthropods summed

up were much more numerous during winter in spruce as compared to

beech (Table 5).

Many groups such as beetles (Coleoptera), Hymenoptera and

homopterans, although abundant, even during autumn, were absent on

beech in January. The total of arthropods sampled on beech was only

one eighth of the number found on spruce. Spiders, an important food

source for birds, were an order of magnitude more abundant on branches

of spruce in comparison with beech during that period of time (Fig. 33).

These first results indicate the importance of more complex structural

features in a forest canopy for hibernating arthropods as observed in

lower forest strata by Gunnarson (1990, 1996).

Fig. 30. Diurnal course of ozone concentrations in ambient air (reference, ‘1 x O3’, above canopy, height: 26m) and in a vertical

profile in the fumigated crown volume (‘2 x O3’, heights: 14, 20, and 26m) as an example on Sept. 12th, 2000. Elevated levels of

ozone occur only in the tree crowns, where the target value (cut at 150 nl l-1, calculated) is reached (‘2 x O3’, heights: 14, 20, and

26m), whereas below the fumigated foliage (height: 6m) ozone concentrations are comparable to ambient air.

78 79

Taxa Beech Spruce

17.10.01 10.01.02 12.04.02 17.10.01 10.01.02 12.04.02

Coleoptera 20 0 3 27 3 26Araneae 16 1 3 213 10 60Blattodea 1 0 0 6 0 0Dermaptera 7 0 0 1 0 0Neuroptera 1 0 0 26 0 0Raphidioptera 0 0 0 0 3 1Hymenoptera 4 0 4 87 3 7Heteroptera 7 0 0 11 0 6Homopterans 43 0 5 61 13 17Aphidina 1 0 0 5 0 0Lepidoptera 0 0 0 0 0 1Holometabolan larvae 1 2 4 8 3 2Collembola 6 0 1 2 0 10Acari 0 0 0 0 0 0Diptera 1 2 0 11 3 6Thysanoptera 0 0 0 0 2 0Psocoptera 0 0 0 183 1 1

Total 108 5 20 641 41 137

Table 5. Total of different arthropod groups in samples of branch beating (n = 20 per treespecies) in the upper and outer canopy of beech and spruce at the Kranzberg crane site.

The crane allows us to study arthropod communities not

only on the scale level of the entire tree crown but also on

smaller scales including vertical stratification, direction

and differences in branch structure.

Perspectives towards future research and cooperationwith the canopy community

Research activit ies wil l remain concentrated on

fundamental questions about the interaction of structure

and function in forest trees and the influence of biotic

and abiotic stressors. The ozone fumigation experiment

will be continued until the end of 2005 at least. Greater

effort will be concentrated on the root systems and soil

processes since these may be considered as inaccessible

as the uppermost canopy.

Stand thinning by removing trees from the research plot

within the next three years will allow the study of dynamic

competition for light in the canopy and at the forest

ground. Experiments using stable isotope techniques intend to give a sound understanding of the processes

of carbon and nitrogen allocation in the trees.

Exchanging experience with international partner groups active in canopy research will facilitate

standardisation of methods in canopy research (e.g., classification of herbivore damage, quantification

of competitiveness, statistical approaches and modelling). Colleagues planning investigations

complementary to those conducted at present are welcome to join the research activities at the Kranzberg

forest.

Acknowledgements

The authors acknowledge the valuable help and assistance by B. Baumeister, T. Feuerbach, N. Hofmann,

G. Jakobi, A. Jungermann, A. König, A. Knötig, P. Kuba, J. Lebherz, B. Rappenglück, H. Reitmayer and

I. Süss. We are grateful to Prof. Dr. M. Kazda, University of Ulm, for providing us with his LAI-2000

instrument.

Background

In 1995, W. Morawetz initiated the development of a project in Southern Venezuela (Surumoni Project,

Chapter 4.3.4), which involved the first tower crane moving on a rail track (120m long) inside a forest.

Three years later, he planned to install a similar crane system in a central European temperate deciduous

forest. The incentive was to compare functional processes in an Amazonian forest with those in a temperate

European forest. In close co-operation with the Centre for Environmental Research Leipzig-Halle (UFZ)

and with strong support from the mayor of Leipzig, a crane system almost identical to that used at

Surumoni was installed in March 2001 within a nature protection area (NSG Burgaue), in the north-

western part of the extensive floodplains within the city of Leipzig (Project LAK, ‘Leipziger Auwaldkran’;

Morawetz & Horchler, 2002).

The criteria to choose the location of the crane included:

• the existence of good baseline information on biotic and environmental data;

• a forest stand with high tree diversity and near-natural species composition and structure;

• minimal logistic constraints on reaching the location e.g. the possibility to use public transport; and

• a well structured floodplain forest comparable to a tropical rain forest.

The site finally selected also included the following: (1) a pre-existing dyke on which the rail track was

built with minor disturbance to the forest; (2) the presence of a nearby building which could be used as

a field laboratory; and (3) the likelihood that the forest will be flooded in future years.

4.2.3. Leipzig Canopy Crane Project (LAK), GermanyWilfried Morawetz & Peter J. Horchler

Fig. 34. Aerial view of the Leipzig crane plot area (June 2002).

80 81

Project management

The project is managed by both authors of this chapter and is hosted at the Institute of Botany of the

University of Leipzig. The project is a non-profit and strictly scientific enterprise. The Federal Ministry

of Education and Research provided an initial three-year grant (¤ 350,000) to install, maintain and

administer the crane. Whilst the crane administration itself does not fund any project, proposals for

further funding are pending. To cover part of the operating costs, a fee of ¤ 25 per hour is charged to all

researchers using the crane. However, costs applying to significant projects, such as long term

investigations, are negotiable. Visiting researchers should apply by e-mail to one of the authors describing

their research project and desired time to use the crane. This should be done at least 3 (to 6) months in

advance in order to reserve crane time.

The Leipzig Crane is a Liebherr 71EC construction crane, movable on a 120m long rail track (Fig. 34

and Table 6), very similar to the Surumoni Canopy Crane. Researchers have access to a forest area of

about 1.6ha and reaching 33m high, by means of two types of gondolas. The purchase of the second-

hand crane, which needed some maintenance work, was partly sponsored by the Liebherr company, to

whom the authors are very grateful.

Variable Characteristics

Location Floodplain in the city area of Leipzig, Germany51°20´16´´N, 12°22´26´´E

Altitude 102mMean annual air temperature 8.8˚CMean annual rainfall 512mmType of forest Floodplain forest of the upper alluvial zone (Querco-Ulmetum)Area of forest accessed by the crane 1.6haCanopy height Max. 36mCrane model Liebherr 71EC, mobile on a 120m trackHeight of tower / Length of jib 40m/45mMaximum height reached by the gondola ca. 33mGondola types Rectangular, circa 1m by 1.5m, 2.2m high, weight 190kg (max 440kg)

Cylindrical, diameter 0.9m, 2.7m high, weight 160kg (max 500 kg)Number of persons carried by the gondola 3, 2In operation since 2001Main research topics Interdisciplinary research on the functioning of the forest ecosystem, including

studies related to biodiversity, biological processes, climate and soil. Forbiodiversity, these include:• Inventory of canopy organisms• Spatiotemporal patterns of canopy organisms and environmental correlates• Role of selected species and species groups on forest functioning• Human impact and its influence on forest functioning

Remarks The set up is near identical to that of the Surumoni crane (Chapter 4.3.4)Management University of Leipzig, Institute of Botany, Systematic BotanyContacts Prof. Wilfried Morawetz, University of Leipzig, morawetz@uni-leipzig.de

Peter Horchler, University of Leipzig, horchler@uni-leipzig.deWeb site www.uni-leipzig.de/~instbota/LAK.htmList of publications Soon on the web site

Fees for researchers 25 ¤ per hour; fee for long term investigations can be negotiated

Table 6. Site and crane characteristics of LAK in Leipzig. Site description and crane details

The climate of the Leipzig area is characterized as intermediate between

maritime and continental (mean annual temperature = 8.8°C and mean

annual precipitation = 512mm). The soils at the crane site are nutrient-

rich loamy floodplain (alluvial) soils. The vegetation is classified as typical

floodplain forest of the upper alluvial zone (Querco-Ulmetum). Due to

river rectifications and canalization, as well as extensive brown coal

mining activities since the early 20TH century, the ground water level in

the Leipzig floodplain forests dropped significantly. Thus, the forest

suffered a gradual but notable change in species composition, favoring

for example Sycamore (Acer pseudoplatanus) which today represents the

most frequent tree species.

The forest stand at the crane site is characterised by a fairly diverse

composition of woody species (17 tree species and 5 shrub species with

diameter at breast height (dbh≥1cm), including 4 introduced tree species

(neophytes). Table 7 summarizes the inventory of trees in the 1.6ha of

forest accessible from the crane gondola.

For means of comparisons and for the dbh class ≥10cm, this corresponds

to 389 trees per ha or 43.2 m2 basal area per ha. Mean tree height is

25.6±6.9m (SD), with a maximum of 36m.

This species composition (Fig. 35) is a product of centuries of practice

of selective logging (‘Mittelwaldwirtschaft’), during which old oak trees

(Quercus robur) were used as timber wood while other species (Fraxinus

excelsior, Carpinus betulus, etc.) were cut much more frequently and used

as firewood. This practice ended in 1870. Since that time there have

been only minor timber extractions and tree species which were

frequently cut in the past grew up to the canopy. Thus, the actual canopy

is formed by old oak trees (> 250 years) and younger trees of that species

(<130 years). A peculiarity of the stand is its amount of dead wood which

provides an important habitat for some rare and endangered insect

species, such as wood-boring beetles. These traits make the research site

a highly valuable nature sanctuary.

The demography of the whole stand tends to show an inverse exponential

curve of stem diameter distribution, typical for many natural forests (Fig.

36). The deviations are so far unexplained. A dendrochronological study

is planned.

Main research topics

The interdisciplinary research will address crucial questions about the

functional processes that regulate the forest. Cooperation and

Species Abundance BA (m2) Frequency IVI (%)

Acer pseudoplatanus 226 13.72 119 69.3Fraxinus excelsior 111 24.09 78 65.2Tilia cordata 229 10.56 112 63.3Acer platanoides 154 2.52 73 34.9Ulmus cf. minor 93 0.99 41 19.6Carpinus betulus 67 2.61 42 19.6Quercus robur 15 6.56 11 14.1Quercus rubra 4 1.01 4 2.8Ulmus cf. glabra 7 0.04 7 2.2Fraxinus pennsylvanica 5 0.29 4 1.8Acer campestre 4 0.35 4 1.8Cerasus avium 4 0.33 4 1.7Robinia pseudoacacia 3 0.5 3 1.3Aesculus hippocastanum 4 0.3 3 1.1Populus x canadensis 2 0.21 2 0.9Crataegus sp. 1 0.01 1 0.3

Totals: 929 63.57 508 300

Table 7. Tree species composition (trees with dbh≥5 cm) in the 1.6ha LeipzigCanopy Crane plot and their rank of importance. The Importance Value Index(IVI: Curtis & McIntosh, 1951) is the sum of the relative number of species(Abundance), relative number of occurrences in all 10 by 10m grid cells of thecrane plot (Frequency) and relative basal area (BA). The relative Abundance,Frequency and BA have been omitted. Note that the total of the IVI is 300%.

Fig. 36. Distribution of all stem diameters ≥5cm dbh into diameter classes.

82 83

organization are based on the scheme illustrated in Fig. 37.

The biodiversity investigations are covered by Martin Schlegel and

Wilfried Morawetz, whilst plant-animal interactions are studied by Stefan

Klotz. Physiological processes, forest structure and genetics, and

environmental conditions are investigated by Christian Wilhelm, Andreas

Roloff and Christian Bernhofer, respectively.

As far as possible, the data and results will be compared to those of

tropical sites, such as Panama (Chapter 4.3.5), Surumoni in Venezuela

(Chapter 4.3.4) and Cairns in Australia (Chapter 4.3.1).

The key objectives, with particular emphasis on biodiversity, include

answering the following essential questions, which are increasingly

complex:

• Species diversity and inventory of canopy organisms.

• Maintenance of biodiversity. Including spatial and temporal

structuring, e.g. vertical, horizontal and temporal patterns of canopy

organisms.

• Interdependence among organisms and environmental parameters,

such as (micro-) climate, water regime, etc.

• Function and functional aspects of selected species and species

groups within the forest ecosystem. Identification of keystone

organisms or keystone guilds.

• Human impact and its influence on the functional processes

identified above. What are the applied issues of our conclusions?

Fig. 36. Distribution of all stem diameters ≥5cm dbh into diameter classes.

Fig. 37. LAK Project organization scheme.Table 8. Preliminary species richness of organisms investigated in the LAK plot.

No. of No. of No. of No. ofobserved canopy exclusive expected

Group species species canopy species species*

Trees (≥1cm dbh) 17 - - 17Shrubs (≥1cm dbh) 5 - - 5Herbs and grasses 37 3 0 >40Mosses and liverworts 17 17 4 >18Lichens 20 20 20 >20Lignicolous macrofungi 56 56 56 >100Slime moulds (Myxomycota) 15 15 ? ?Water bears (Tardigrada) 3 3 ? ?Butterflies (Lepidoptera) 27 27 ? >100Ground beetles (Carabidae) 21 23 ? > 23Wood-dwelling beetles (Coleoptera) 105 105 ? >150Bugs (Heteroptera) 58 58 ? >70Ants (Formicidae) 5 5 3 7Bumble Bees (Bombus spp.) 8 8 ? >8Orb-web spiders (Araneidae) 46 46 ? >100Amphibians 4 1 0 4Bats 5 5 ? 7

*Rough estimates

It is obvious that with increasing complexity there will be a decreasing number of species that can be

studied in detail. Nevertheless, we would emphasize the major importance of organismic biology in our

project. The main goals of the observation cranes are to allow observations, measurements and sampling

in situ. All of this information will be databased and linked to a GIS in order to provide reference data

for all project participants.

Main findings

Since the crane has recently been established, to date only some, mostly short-term studies have been

performed. A first overview of the inventory of forest and canopy biodiversity is given in Table 8.

Surprisingly, observations in the canopy emphasized a rather high diversity in macro fungi (56 spp.), as

well as in slime moulds (Myxomycota, 15 spp.) and the re-colonization of formerly very rare and threatened

lichen species (11 spp.), clearly indicating improved environmental conditions, i.e. decreasing air pollution.

Some coincidental observations will lead to more detailed studies, and these included:

• the presence of the frog Hyla arborea in the upper canopy (first physical proof for Germany),

• the presence of snail species (Arianta arbustorum, Cepea sp.) foraging on mosses in the upper canopy,

• a bird (Parus caeruleus) acting as potential pollinator for a ‘usually’ wind pollinated tree species

(Fraxinus excelsior).

Another major research project that was initiated in 2002 was the investigation of tree phenology focusing

on the species Fraxinus excelsior, Tilia cordata and Quercus robur. Detailed study of Fraxinus excelsior

revealed a striking spatial and temporal variability in generative and vegetative phenology, in the

distribution of the three flower types as well as in leaf and branching pattern and morphology. The

investigation of this focal species will be continued including more detailed studies on reproductive

ecology and population genetics. A comparison with the results from the Tomakomai Canopy Crane

Project (Chapter 4.2.5) concerning the Japanese tree species Fraxinus lanuginosa (Ishida & Hiura, 1998)

is planned.

So far, three joint projects with foreign institutes have been initiated, two including visiting researchers.

A former co-worker in the Surumoni Canopy Crane Project, Klaus Jaffé (Universidad Simon Bolívar,

Caracas, Venezuela) performed a pilot study on canopy ants using baits. This method indicated that the

abundance of ants was very low in the canopy, quite unlike that of

Surumoni (K. Jaffé et al., unpubl. data).

Geoffrey Parker (Smithsonian Environmental Research Center,

Edgewater, USA) was invited to perform a study of the canopy and

understorey topography by means of a laser rangefinder system (LIDAR)

to get essential baseline data for all collaborators of the project. The

measurements covered a part of the LAK plot. The results for the canopy

topography are illustrated in Figure 38.

In fact this represented the first comparison of the same canopy at the

same location measured with the same instrument both from above and

Fig. 38. Surface representation of the canopy topography in the westernpart of the canopy crane plot, based on a small subset of LIDAR measurementson an area of 45m by 110m. Note the hole (bottom left) a huge gap createdin 2002.

84 85

below. This comparison revealed a notable difference in the arrangement

of leaves (Fig. 39).

Whilst leaves are clearly grouped at the outer canopy in order to optimize

their position to light no such pattern can be recognized for the leaves

and twigs beneath. Overall statistics of height measurements with LIDAR

are indicated in Table 9. The standard deviation is a measurement for

the roughness of the canopy (G.G. Parker, pers. comm.).

The measurements also permitted the calculation of the mean vertical

profile of surface area density (including leaves, twigs and bark; Fig.

40).

David Shaw (Washington University, USA) and Kristina Ernest (Central

Washington University, USA) incited the present authors to perform a

study on stand level herbivory at LAK. A similar protocol was also applied

at various tropical and temperate canopy crane sites in order to obtain

global comparative estimates of percent herbivory damage of the whole

stand. At 100 randomly selected sampling points, 10 leaves are selected

randomly and percent leaf damage is measured for all leaves. In 2002 at LAK, 57 sampling points only

(570 leaves) were studied. Based on these data, which were definitely too restrictive for our large study

plot (1.6ha) we calculated a percent leaf damage of 1.12%, which appears far below corresponding values

obtained for tropical forests. It is planned to repeat this study in 2003, this time collecting circa 1,500

leaves at 150 random sampling points.

Future research and collaboration

During this first research phase, the intention was to obtain an overall canopy inventory, but in the

future emphasis will be the identification of spatial (particularly vertical) patterns of biodiversity and

measuring their environmental correlates, particularly microclimatic parameters. In addition, other studies

will be expanded, particularly phenological ones, whilst physiological, biochemical and genetic studies

are awaiting funding to be initiated.

A further step will be the replication of the WebFace experiment, currently being performed at the Swiss

Canopy Crane by Christian Körner (Chapter 4.2.1). In his experiment, parts of the forest trees are exposed

to a CO2-enriched atmosphere in order to measure the physiological and, ultimately, the ecosystem

response. The final aim is to assess the response of the forest to anthropogenic increases in atmospheric

concentration of carbon dioxide.

Certain projects are particularly suitable for collaboration with other canopy crane sites. These are the

LIDAR and stand level herbivory projects, both so called pathfinder projects of the Global Canopy

Programme (see Chapter 2). If possible, they should be replicated at all canopy crane sites.

A pilot study on the spatial and temporal dynamics of thermal properties of the forest stand was performed

in 2002 by Jörg Szarzynski and colleagues (Center for Development Research, University of Bonn and

Department of Physical Geography, University of Mannheim, Germany). Using a high-precision

Fig. 40. Mean canopy height profilebased on LIDAR measurements withestimated surface area. Error bars representstandard errors.

thermographic camera, thermal scans of the entire forest were generated, from the soil to the canopy.

Preliminary results are very auspicious and may yield new insights into the heat balance of vegetation

plots. Since spatial data are recorded, this method provides important information for several scientific

disciplines, such as bioclimatology, physiology, botany and zoology, or generally to those projects interested

in the environmental impact of microhabitat or niche differentiation. Hence, standardized thermographic

mapping should be regarded as a method highly suitable to be included in the frame of GCP pathfinder

projects.

Another experiment quite amenable to collaboration has been proposed by S. Joseph Wright (Smithsonian

Tropical Research Institute, Panama) and already performed in Panama (Chapter 4.3.5, Van Bael). It is

intended to exclude predators of arthropods, mainly birds and bats, in order to monitor the response of

insect herbivores. This experiment will be replicated in one of the next field seasons.

Since it was rather surprising to find such a high diversity of slime molds (Myxomycota), a collaboration

is underway with Martin Schnittler (University of Greifswald, Germany), with the intention of comparing

data on this group from several sites (Australia, Costa Rica, USA). A similar effort should be invested

into a global comparison of canopy macrofungi.

A collaboration initiated by Andreas Prinzing (University of Mainz, Germany) is planned for 2003 with

colleagues from Poland investigating the microfauna, i.e. Acari, Collembola and Tardigrada. These non-

flying and less mobile animals are expected to be much more site specific. Hence beta diversity among

the microfauna community in the forest canopy is predicted to be higher than among more mobile

arthropods. A similar project has been performed by Australian colleagues and appears to be quite

stimulating and suitable for a global collaboration.

Finally, in our opinion collaboration is needed for a detailed study of generative and vegetative tree

phenology. Therefore, some standard protocol, such as the one used in the stand level herbivory project,

needs to be developed.

Many of these collaborations can be facilitated by the Global Canopy Programme and the International

Canopy Network. Both provide excellent contacts and databases and we are convinced that a global

collaboration is the only way forward to promote canopy research.

Fig. 39. Comparison of LIDAR measurements (subset) of the outer canopy(small bars connected by line) and “understory height” (triangles; the shortestdistance from the ground to the lowest leaf, branch or twig) along the sametransect of 112m length.

Variable Outer canopy height (m) Understorey height (m)

Mean 25.6 9.2Median 27.6 7.6SD 6.9 7.0

Table 9. Statistics associated with LIDAR measurements of the height ofthe canopy surface (outer canopy) and understorey.

86 87

Location and site characteristics

The Solling crane is located in the centre of a triangle of roof installations underneath the forest canopy,

consisting of three roofs of 300m2 surface area each (Figs 41 to 44 and Table 10). This roof facility, part

of a European network, was installed in order to perform whole-

ecosystem manipulation experiments under field conditions, and it was

the only one with a device to gain access to the canopy. In particular, the

crane was designed as a device to facilitate physiological measurements

in the canopy to investigate tree responses to the experimental

manipulations.

The roof manipulation facility and the crane were located at Solling since

this experimental forest has been a focus of intensive interdisciplinary

forest ecosystems research since the 1960s. The site was part of the Man

and the Biosphere Programme (MAB) and the International Biosphere

Programme (IBP). Therefore, data from former monitoring and research

projects were available to be used as baseline information for the

ecosystem manipulation study.

The Solling experimental forest is located in central Germany, in a

mountainous area which is a part of the Weser river mountain range.

The roof manipulation study is run in a 67 year old (2003) stand of Picea

abies (L.) Karsten (Norway spruce). The site is located at circa 500m

elevation a.s.l. on a plateau. Average annual rainfall is 1,090mm. The

soil is a strongly acidified dystric cambisol (FAO classification) which

has developed in a loess solifluction layer overlying sandstone bedrock.

Base saturation is 7% or less of cation exchange capacity (i.e., capability

of the soil to store nutrient cations) throughout the mineral soil profile.

The spruce forest is in a state of impaired vitality, with symptoms of

needle losses and yellowing, the latter due to severe magnesium deficiency

in the highly acidified soil. The roofs were built underneath the canopy

with the roof ridge at circa 3.5m above the ground. They are permanent

timber frame constructions, covered with highly transparent

polycarbonate plates. Water falling onto the roofs is collected in a central

cabin, processed in various ways and redistributed to the roof plots by a

sprinkling system.

Installations, experiments and key research topics

The Solling roof project is an interdisciplinary study comprising monitoring of the soil (both chemistry

and hydrology), the forest stand (roots and above ground physiology), the ground vegetation and the soil

fauna and microflora, as well as micro-meteorological parameters. A detailed listing of installations and

monitoring may be found in Bredemeier and Dise (1992).

4.2.4. Solling, GermanyMichael Bredemeier, Achim Dohrenbusch & Gustav A. Wiedey

Fig. 42. View from the gondola to thecrane tower and the canopy of the 24mhigh forest of Norway spruce (photo AchimDohrenbusch).

Fig. 41. Ground plan of the Solling roof project.

The experimental treatments applied in the Solling roof experiment were (Bredemeier et al., 1998):

1. ‘Clean rain’ roof with simulated pre-industrial throughfall. Water is filtered, de-ionised, adjusted to

clean rain - concentrations and redistributed to the plot immediately (designated as site ‘D1’ or roof 1).

2. Control roof for roof effects alone, without further manipulation. Water is re-sprinkled immediately

without changing ion content (site ‘D2’, roof 2 or control roof).

3. Drought/re-wetting roof, for simulation of strong drought events with subsequent intensive re-

wetting. Water is stored in a tank battery during the experimental drought phases and reapplied to

the site during the re-wetting phases (site ‘D3’ or roof 3). The durations of drought periods is in the

range of 10-25 weeks, up to 140mm of throughfall can be stored during drought and reapplied in

the re-wetting phases. Experimental drought periods were conducted at different times of the year

between spring and autumn.

4. Control without roof, consisting of ambient, non-manipulated throughfall (site ‘D0’ or ambient

control).

The composition of the artificially prepared pre-industrial through-fall for the clean rain plot (D1)

corresponded to a reduction of sulphur input of circa 65% and a reduction of total nitrogen

(ammonium+nitrate) input of 80% relative to the ambient control in the reference years 1990/1991 (i.e.,

before the activation of the manipulation). The sea salt constituents sodium and chloride are also decreased

significantly in the sprinkling solution, since they are considered non-essential for the forest nutrition.

Metal nutrient cation contents were only moderately changed in the artificial clean rain solution.

The aim of the drought/re-wetting roof experiment (plot D3) at Solling is to observe the extent of direct

drought stress to the forest, but also to detect secondary drought effects in soil water chemistry. With

respect to the latter, it has been hypothesised that in phases of re-wetting after intensive drought

‘acidification pulses’ due to net nitrification would occur (Ulrich, 1983), increasing stress to the ecosystem

by adding soil chemical stress to that of drought.

Liquid samples were collected weekly from bulk deposition, throughfall and soil water at several depths

(0, 10, 20, 40, 70 and 100cm mineral soil depth) and combined into monthly samples for analysis. In the

liquid samples, pH was measured and concentrations of Na, K, Ca, Mg, Al, Fe, Mn, St and Pt were

determined, as well as Cl-, NO3- and NH4+ concentrations. Litterfall and fine roots from the soil were

collected seasonally, pressure-digested and analysed for metal nutrient cations, C, N, S, and P. Suction

cup lysimeters were installed at the four sub-plots of the roof experiment in the autumn of 1989, well

before the construction of the roofs and the manipulation treatments began. Roofs were closed and

sprinkling systems started to operate in September 1991.

In order to investigate physiological responses of the trees to experimental manipulations, it was necessary

to gain access to the canopy. In spring 1992 a 33m high crane was installed in the centre of the roofed

area. It was equipped with a special transport system for personnel (a gondola with a floor space of

100x70cm) which made it possible to reach the crown area of all of about 100 trees which belonged to

the experimental sites. With this method regular physiological measurements of the carbon budget

(photosynthesis, respiration) and the water budget (transpiration, xylem water potential, osmotic potential)

as well as the shoot elongation and fruiting intensity were performed (Dohrenbusch et al., 2002a). In

addition, each year needle samples were taken for element analyses and also an assessment of forest

health by visual observation of the needle losses and needle yellowing was performed.

Fig 43. View up the crane tower (photoGustav Wiedey).

Fig. 44. View down upon the gondola(photo Gustav Wiedey).

88 89

Key findings from the crane experiments

Height growth

The course of the annual height growth for all experimental variants showed a development unaffected

by treatment. The mean height increment decreased continuously since the beginning of the measurements

in 1988 from an average of 37cm on all sites to a minimum in the fifth year of monitoring in 1992 (Fig.

45). At that time the mean shoot growth was only 14cm. Afterwards, up to 1996, a marked increase in

height growth was again observed. In the years 1993 and 1994, influence of the drought treatments on

the height increment was observed. As a result of the long dry periods during the summer months of

previous years the mean height increment on the D3-site was significantly reduced by about 50% compared

to the other sites. After 1995, no effects induced by the droughts in

previous years could be discerned. By contrast, the effect of the de-

acidified precipitation on the height growth of trees at the D-1 site for

the total period monitored was not statistically significant (Dohrenbusch

et al., 2002b, 2002c, 2002d).

Fructification

Figure 46 shows the average numbers of spruce cones in the years 1992

to 1998. The first count in late summer 1992 showed a large number of

cones with an average between 93 and 97 per tree. However, the mean

values on an area basis concealed differences between individual trees.

Variable Characteristics

Location Solling mountains, Germany51°31´ N, 9°34´ E

Altitude 500mMean annual air temperature 6.4 °CMean annual rainfall 1090mmType of forest Norway spruce plantationArea of forest accessed by the crane 0.2haCanopy height 24mCrane model Liebherr modified, based on model 32K/45 (year of manufacture

1977), fixedHeight of tower / Length of jib 33m/25mMaximum height reached by the gondola 28mGondola type Rectangular, 1.5 x 0.7 x 1.2m, carrying capacity 500kgNumber of persons carried by the gondola 2In operation since 1992Main research topics • Responses of trees to experimental manipulations on roof plots

• Precision growth measurements• Precision assessment of vitality status and damages• Photosynthetic capacity and time courses of photosynthesis rate• Trace gas exchange in the canopy

Remarks Operation by trained personnel onlyManagement Dr. Gustav A. Wiedey, University of Göttingen, gwiedey@gwdg.deContact Dr. Michael Bredemeier, University of Göttingen,

mbredem@gwdg.deWeb site http://www.gwdg.de/~fzw/homee/sollingt.htmList of publications http://www.gwdg.de/~fzw/homee/publ.htmFees for researchers To be negotiated

Table 10. Site and crane characteristics of the Solling Canopy Crane.

Fig. 45. Development of the annual height growth of the 60 to 70-year-oldtrees (number of trees per roof: 23-27).

Whilst some trees had several hundred cones, others had none. During

the following two years the amounts dropped to 30 cones in 1993 or 5

cones in 1994, respectively. After a new increase in the years 1995 and

1996 the number of cones in 1997 attained similar numbers to those of

1994 (Dohrenbusch et al., 2002b, 2002c, 2002d).

Tree vitality and visual appearance of the crown

From 1993 onwards, needle loss and yellowing were monitored for the

trees of the roof and control areas. This assessment of the crown was

performed from the crane by the same person every year. However,

comparisons between years have to be viewed with caution. Therefore,

for almost all the trees, photographs of the crown were taken from the

same perspective, to optimise the annual damage assessment. During

the observation period the tree vitality increased continually. This

development was perceived on all of the roofed plots (Fig. 47). The effect

of experimental treatment was particularly distinguishable on the trees

in the drought experiment, which exhibited particularly high needle loss

(Fig. 47) and severe yellowing of needles (Dohrenbusch et al., 2002c).

Nutritional status of the trees (mineral element concentrations)

To determine the element concentrations in needles from 1992 onwards,

needle samples were taken from all trees of the roof and control sites.

Sampling took place in September from the sun crown of the trees (6th

whorl from the top). Only green branches were harvested, and only

needles of the two most recent years were investigated. In order to

estimate the nutrient concentration for the whole tree, detailed

investigations into the variation of nutrient concentration in the crown

have been made. The crane with the transportation system permits the

access to all parts of the tree crown. Needle samples were obtained from

each year and different crown levels (tree heights). Figure 48 shows as

an example the trend of magnesium concentration according to crown

position and needle age (Dohrenbusch et al., 2002c).

Conclusions from the Solling crane measurements

The crane at Solling was used to measure above ground responses of the

spruce stand to the experimental manipulations in the roof project as

described above. In the clean rain experiment, above ground effects were

generally weak, while the effects of the altered input on soil water

chemistry were rapid and strong. Also the root system responded by

increased fine-root growth and an improved morphology, but with a

time lag of several years. The measurements from the crane are being

continued at a lower level of intensity, in order to check whether responses in the canopy will occur with

an even longer time lag. In contrast to the clean rain manipulation, canopy level responses in the drought/

re-wetting experiments were strong: height growth declined significantly in years following experimental

drought phases, and photosynthetic parameters decreased during drought treatments. However, all these

parameters also exhibited relatively fast recovery after the end of the experimental manipulations.

Fig. 46. Cone production during the period 1992-1998.

Fig. 47. Observed status of defoliation for the period 1993-1998.

Fig. 48. Relative concentration of magnesium depending onthe position in the crown (relative tree height) and needle age.

90 91

Background

The Tomakomai canopy crane was installed in November 1997 in the 2,715ha Tomakomai Experimental

Forest (TOEF), situated near the town of Tomakomai in Japan (Table 11). The forest is adjacent to the

Utonai bird sanctuary in the Yufutsu marsh. The crane was funded by the Ministry of Education, Science,

Sports and Culture of Japan, through the programme “An integrated study on biodiversity conservation

under global change and bio-inventory management system”. It is managed by the Tomakomai Research

Station, a unit of the Hokkaido University Forests. The University forests were established in 1901 and

total 70,000ha, including a tree breeding station, a tree nursery, two experimental stands, and six regional

forests, including TOEF. TOEF hosts 150 visiting scientists each year, facilitating their work by providing

laboratories, housing, libraries, and research permits. Dormitories are available for visiting researchers

and students using the station for educational and research purposes, with the Director’s permission.

Long-term dormitories with kitchen can accommodate a maximum of 20 people, whilst short-term

dormitories can accommodate 42 people.

4.2.5. Tomakomai Experimental Forest, JapanMasashi Murakami & Tsutom Hiura

Variable Characteristics

Location Tomakomai Experimental Forest, Hokkaido island, Japan42°40' N, 141°36' E

Altitude 90mMean annual air temperature 6.1°CMean annual rainfall 1254mmType of forest Temperate deciduous broad-leaved forestArea of forest accessed by the crane 0.5haCanopy height 20mCrane model Ogawa Industries, OTH-80N, fixedHeight of tower / Length of jib 25m/41mMaximum height reached by the gondola 20Gondola type a: Cylindrical, 2.1 x 0.7m

b: Cylindrical 2.4 x 1.2mNumber of persons carried by the gondola a: 1 person

b: 5 personsIn operation since 1997Main research topics • Forest structure and species richness

• Diversity and productivity in the forest ecosystem• Carbon and nitrogen dynamics of a deciduous broad-leaved forest• Climate change• Seasonal exchange of leaf-level characteristics in relation to successional traits• Mechanisms of masting in relation to storage reserves in tree species• Relationships between the three-dimensional structure of the forest and insect

diversity• Other research topics as detailed in the text

Remarks Located in an experimental plot of 9ha in which a large scaffolding unit also standsManagement Tomakomai Research Station, Hokkaido University ForestsContacts Prof. Tsutom Hiura, Director, Hokkaido University, hiura@exfor.agr.hokudai.ac.jp

Prof. Masashi Murakami, Hokkaido University, masa@exfor.agr.hokudai.ac.jpProf. Naoki Agetsuma, Hokkaido University, age@exfor.agr.hokudai.ac.jp

Web site http://pc3.nrs-unet.ocn.ne.jp/~exfor/toef/toef.htmlList of publications http://pc3.nrs-unet.ocn.ne.jp/~exfor/toef/toef.htmlFees for researchers Dormitory fees only

Table 11. Site and crane characteristics of the Tomakomai Canopy Crane.

Canopy Crane

The crane is located in a mature deciduous forest with Ostrya japonica,

Acer mono, Cercidiphyllum japonicum, Quercus crispula and Tilia japonica

among the most common species (Fig. 49). The crane can reach 20 species

of trees and 6 species of liana. The crane perimeter covers an area of

0.5ha and is part of an experimental plot of 9ha in which all individuals

of 10cm diameter at breast height (dbh) or greater have been measured,

mapped and identified. One ha of this forest includes 458 stems larger

than 10cm in diameter, so that the total basal area is 27m2 per ha. TOEF

is situated on a deep layer of volcanic material (2m in depth), originating

from the last eruption of Mt. Tarumae in 1739. The slope of this forest is

rather moderate (from flat to 10°).

The Japanese archipelago is located along the green belt of the western

Pacific and Asia, where a humid climate extends from northern Siberia

to southern New Zealand. It is an area with high biodiversity and productivity. TOEF is a core site for

several international programs, including DIWPA-IBOY (International Biodiversity Observation Year)

and DIVERSITAS. In these programmes, the diversity of forest plants, vertebrates, and invertebrates are

monitored to study functional aspects of the ecosystem. The diversity of targeted organisms is determined

using standardized methods (Nakashizuka & Stork, 2002). TOEF welcomes collaborative research on

any field of canopy ecology and biology.

Below, we present the main research topics and results of TOEF projects with respect to (a) forest structure,

productivity and climate changes; (b) plant resources in the forest; (c) arthropod diversity and forest

structure, and (d) animal-plant interactions.

Forest structure, productivity and climate change

Relationships between diversity and productivity

Forest ecosystems include a large number of species and are the greatest reservoir of carbon of terrestrial

ecosystems on Earth. However, anthropogenic activity is raising the level of carbon dioxide in the air and

reducing biodiversity. Recent experiments on microcosms and grasslands have shown that as the number

of plant species increases, the net primary productivity and nutrient cycling in these systems also increases.

However, no empirical tests of this result have been attempted in forest ecosystems, due to the difficulties

in either manipulating the forest experimentally or selecting appropriate fields that have uniform

environmental conditions to test this controversial hypothesis. In this study, we used an ideal forest

ecosystem, in which environmental conditions were uniform among the stands, to show that ecosystem

productivity in 40 forest plots increased significantly with an increase in tree diversity. The results of this

study imply that the preservation of biodiversity is essential for the well-functioning of carbon cycle in

forest ecosystems (Ishii et al., in press; Hiura, in press).

Carbon and nitrogen dynamics of a deciduous broad-leaved forest

Seasonal changes in the soil respiration rate and litter production were measured in TOEF. Soil respiration

rate was detected monthly by an open-ended soil respiration chamber and was significantly correlated

with soil temperature, at about 10cm depth. Litter fall was monitored with twelve 0.88m2 traps, which

Fig. 49. The TOEF Canopy Crane in summer, in use to count leaves (photoTsutom Hiura).

92 93

were installed at random within the forest. Green litter was observed in summer as a result of the action

of strong winds. Based on the dynamics of carbon and nitrogen, we will shortly estimate carbon flux in

the forest ecosystem, as a function of nitrogen flow through the plants.

The size-dependent decline in the productivity and hydraulic conductance of canopy trees was also

examined. In the forest, canopy trees have higher primary productivity than that of other plants, but

hydraulic conductance may decrease as tree size increases. The decrease of hydraulic conductance may

reduce photosynthesis of canopy leaves. Therefore, we hypothesized that within single species,

photosynthesis of canopy leaves declines as tree size increases due to the increase of hydraulic limitation,

resulting in a size-dependent decline in the productivity of canopy trees. However, maximum sizes differ

among species. Hence, among species, canopy trees with similar sizes may have different hydraulic

conductance. A different hypothesis implies that, among species, canopy trees with smaller maximum

size have stronger hydraulic limitation to photosynthesis of their canopy leaves. These two hypotheses

are being tested on sympatric species of Acer, including A. mono, A. palmatum and A. japonicum (Suzuki

& Hiura, 2000).

Climate change

To infer the responses of plants to changing environments at a large scale,

it is important to be able to make predictions about processes at a smaller

scale, and to classify plants among functional types with respect to these

responses. To understand the factors influencing the seasonal changes

of photosynthesis, forest ecologists at the Hokkaido University

simultaneously measured a suite of environmental conditions surrounding

leaves of 20 tree and liana species. In doing so, they quantified the species-

specific sensitivity to environmental conditions as they change throughout

the growing season. Seasonal patterns of leaf-level photosynthesis appear

to consist of interactions between the suite of ambient environment

conditions and the species-specific sensitivity to this combination of

factors. Furthermore, to investigate the response of forest productivity

to climate changes, the three-dimensional leaf structure in a forest patch

was examined in detail (480,399 leaves were counted and mapped in

total). A production model will be constructed shortly, using the data

obtained in measurements of photosynthesis and forest structure (contact:

Tsutom Hiura).

Seasonal change of leaf-level photosynthetic characteristics as related to

successional traits

Changes in solar radiation, air temperature, and vapour pressure deficit

through the growing season lead, in part, to seasonal differences in

photosynthetic rate, particularly in the forest canopy. To understand the

factors influencing seasonal changes in photosynthetic rate, we

simultaneously measured a suite of environmental conditions surrounding

leaves to quantify the species-specif ic sensit ivity to changing

environmental conditions throughout the growing season. In particular,

we observed species-specific seasonal patterns in photosynthetic rates

in the canopy.

Fig. 50. Measurement of the photosynthetic activity of Quercus crispula inthe forest canopy (photo Tsutom Hiura).

In late successional species, such as maple and sub-canopy hornbeams, the stable period of light-saturated

assimilation rate (Amax

) was longer than other species, whilst light-demanding birch and several species

of gap-specialist magnolias indicated temporal depression of Amax in summer. Amax of mid-successional

species, such as hop hornbeams and oak, reached their maximum value in July, and declined gradually

through the season. Thus, we confirmed that seasonal changes in photosynthesis rates were limited by

both stomatal and non-stomatal factors, such as stomatal closure and reduction of photochemical activity.

In other words, seasonal patterns of leaf-level photosynthesis consist of interactions between the suite of

ambient environment conditions and the species-specific sensitivity to the combination of those factors

(Fig. 50) (contact: Tsutom Hiura).

Plant resources in the forest

Masting and its relationship to storage reserves

Perennial plants allocate a significant fraction of their photosynthetic output to long-term storage. The

continued oversupply of photosynthate suggests that stored reserves have some important ecological

roles. Masting is an intermittent production of large seed crops. Most studies of this phenomenon deal

with the evolutionary advantages of masting, and not with the mechanism of masting. This study will

investigate the mechanism of masting and its relationships to storage reserves in different tree species.

We propose to measure the stored reserves with microscopic observations of plant cells surveyed from

trunks and branches of trees at different heights (0.3m, 1.3m and 4m), from various phenological stages

(Miyazaki et al., 2002; Miyazaki et al., in press; Fig. 51).

Variation of tree sex-ratio during forest succession

In androdioecy (gender expression of male and hermaphrodite trees), the equilibrium frequency of males

depends on their average pollen fitness relative to the average pollen fitness of hermaphrodites. In a

wind-pollinated androdioecious species, selection favours a hermaphrodite-biased sex ratio and self-

fertility in low-density stands, but male frequency increases with increasing population density due to

the increment in the efficiency of wind pollination. To test this hypothesis, the frequency of males was

determined for ten populations of a wind-pollinated androdioecious ash (Fraxinus lanuginosa) and these

frequencies were related to stand structure. Male frequency correlated significantly with the density of

the ash and the existence of sex-labile trees, indicating an environmental component in sex determination.

The average fruit set in a mass-flowering year was positively correlated with forest productivity. The sex

ratio converged on a consistent value in accordance with the prediction of the model of sex ratio for

androdioecy as the forest achieved a dynamic equilibrium. The results showed that environmental

fluctuations and community structure were important in the maintenance of the androdioecious tree

(Ishida & Hiura, 1998, 2002; Fig. 52).

Arthropod diversity and forest structure

Diversity and vertical distributions of flying insects

Many scientists have used the canopy crane to study insect herbivores and pollinators, and their diversity.

Entomologists at the Hokkaido University collected drosophilids, beetles, moth larvae and bumblebees

from the canopy using the crane and other techniques during three years. In total, 402 species of small

beetles, 3,908 bumblebee individuals and huge numbers of drosophilids and moth larvae were collected.

The entomologists investigated the seasonal habitats and three-dimensional distribution of insects in the

Fig. 51. Flowers of Styrax obassia. Thisspecies produces mast-flowering withintervals of circa 5 years (photo YukoMiyazaki).

94 95

Fig. 52. Paper bags installed in ash (Fraxinus lanuginosa) to monitor self-incompatibility (photo Tsutom Hiura).

forest. Common winged insect species were distributed vertically in a specific way in close relation to the

stratified structure of the forest vegetation. Even on a rather small spatial scale (18x18x22m), a stratified

distribution was prominent for most insect species flying above the ground in the forest. Regardless of

the variation in foliage density and tree species among grid-cells at the same height, each insect species

tended to be distributed rather evenly within a particular stratum. Thus, the stratification of vegetation

plays a major role in promoting the diversity of flying insects in forest ecosystems (Fukushima et al.,

1998).

Along the Asian Green Belt, the diversity of the regional species-pool, which has largely been determined

as a historical product of the past geological and biological processes, decreases in parallel with the

decline in complexity of forest structure from the tropics to the polar regions. In order to evaluate the

effects of forest structure itself on the diversity of forest animals, it is necessary to compare animal

communities among structurally different forests within the same regional species-pool. We installed

vertical collecting devices with various types of traps for flying insects in natural and secondary forests of

TOEF. Sampling devices included interception traps (Malaise and window traps) and bait traps. These

traps were easily surveyed and enabled non-destructive, continuous sampling (day and night, and

throughout the year). Those traps were set at approximately 5m intervals from the ground to over the

canopy. Samples were collected weekly throughout the active season. In combination with a rope system,

the traps yielded samples from different strata of forest vegetation. The material was first sorted at the

order level, and species identification is now on-going. The two types of interception traps yielded samples

quite different in the composition of insect species: samples from Malaise traps were dominated by weak

fliers such as small dipterans and parasitoid wasps, whilst those from window traps were dominated by

strong fliers such as large bumblebees, halictine bees, vespids, and beetles. Common winged insect species

were distributed vertically in a specific way in close relation to the stratified structure of the forest

vegetation (contact: Masanori Toda, hutian@pop.lowtem.hokudai.ac.jp).

Three-dimensional distribution of foliage and flying insects

The microdistribution of insects, mainly drosophilids, flying above the ground within the forest was

examined in relation to the three-dimensional distribution of foliage within a relatively small spatial

scale. With the aid of a canopy-access system, which divided a space of 18mx18mx22m of forest into a

grid of 1.83m3 ‘cubes’, a total of 80 banana-bait traps were installed in a checkered fashion at 3.6 m

intervals. The amount of foliage was measured for each tree species and for each ‘cube’ of the grid.

Drosophilid samples were collected every 5 days during the summer from mid July to mid August. Even

within this relatively small spatial scale, the vertical distribution of insects flying above the ground was

stratified for most species. Regardless of the variation in foliage density and tree species among grid-cells

Fig. 53. Moth larvae of a species of Saturniidae. In TOEF, moth inventory is on-going, and collections have so far included 27,966 specimens from 55 families and 2,072 species.This corresponds to 75% of the 2,780 species recorded in Hokkaido (photo Masashi Murakami).

96 97

at the same height, each insect species tended to occur within a particular stratum. Thus, the stratification

of vegetation plays a major role in promoting the diversity of flying insects in forest ecosystems (Tanabe,

2002).

Animal plant interactions

Spatio-temporal variation in the moth larval community associated with oak

Trees provide animals with both habitats and food. The amount and quality of these food resources also

exhibit spatio-temporal variation. For example, the quantity and quality of leaves as food for herbivores

change both between seasons and between conspecific trees growing in the canopy and understorey

(mature trees, saplings and seedlings). Such spatio-temporal variation in the food resources provided by

trees should affect the composition and the diversity of the communities of animal consumers. Moth

larvae were hand-collected from sunlit and shaded leaves of canopy oak trees and their saplings, during

the spring and summer. Several physical properties of the habitats and physico-chemical properties of

the leaves sampled were measured. Preliminary results suggest that during spring, oak sapling leaves

were toughest, followed by shaded and sunlit canopy leaves, reflecting the order of leaf sprouting. However,

during summer the order of leaf toughness was reversed: sunlit canopy leaves > shaded canopy leaves >>

sapling leaves. Such food resources provided by the same host plant but varying spatio-temporally in

quality supported different communities of consumers during different seasons and at different locations

within the forest. As a result, even a single tree species may accommodate diverse communities of

herbivores within a structured forest and in seasonally fluctuating environments (Wada et al., 2000; Fig.

53).

Fig 54. Counting the number of flowers on the ash Fraxinus lanuginosa (photo Tsutom Hiura).

Effects of herbivory and resource availability on induced defenses in oak

Plants have evolved many defenses against herbivores. Among them, plant secondary metabolites could

have evolved as one of the main active defenses against herbivores. Often, these are produced in response

to foliar damage. Since they are by-products of metabolism and only produced under certain physiological

circumstances, their production may be constrained by the availability of plant nutrients. However, it is

still unknown how nutrient availability and foliar damage affect plant chemistry. This study will examine

the response of oaks to herbivory, using saplings and canopy trees at the crane site, under different light

and soil conditions. Leaf mass per area (LMA) and leaf tissue concentrations of carbon, nitrogen, total

phenolics, and tannins are being measured as plant defenses. Preliminary data show significant

relationships among height, light availability, herbivory and LMA (contact: Masashi Murakami).

Variation in resource allocation in trees and its effect on canopy herbivores

The relationships between community structure (abundance and diversity) of herbivores and seasonal

changes in leaf characteristics are being examined. Leaf toughness, nitrogen and tannin content were

measured on 10 canopy tree species. Arthropods were collected by beating. Patterns in resource seasonality

(i.e., allocation of nitrogen and carbon) differ considerably among tree species, according to either their

pattern of leaf-flush or their mechanisms of defenses against herbivores. Seasonal trends in the allocation

of nitrogen and carbon and the patterns of distribution of herbivore communities are being compared

among tree species in the canopy. For example, two conspicuous and seasonal peaks in abundance were

observed when studying the dynamics of populations of caterpillars in the canopy of the oak Quercus

crispula. Our aim is to derive general patterns involving herbivore-plant interactions and the nitrogen

dynamics within the forest canopy (contact: Masashi Murakami).

Fluctuation of tree flower abundance and its effects on the population dynamics of bumblebees and on fruit

production of understorey herbs

The goal of this study was to understand the yearly dynamics of the pollination system within the forest.

Bumblebee populations require abundant flower resources throughout the year. In temperate deciduous

forest, the abundance of flowers from tree species pollinated by insects fluctuates significantly between

years. Hence, the availability of flower resources in trees may affect the population size of bumblebees

greatly. In turn, bumblebee abundance may affect the pollination of understorey herbs. Thus, we examined

the relationship between flower abundance of trees, population size of bumblebees and fruit production

of understorey herbs.

Both visiting insects and the flowering phenology of trees were studied. For ten tree species, we counted

the number of inflorescences on 10 individual trees. We calculated the relationships between breast

height diameter and the number of inflorescences per tree, fitted the resulting function to all trees of a 4

ha plot, and summed these estimations. Fluctuations of the population size of bumblebees were recorded

by installing window traps at various heights in the forest. The visiting pollinators and fruit production

of some understorey herbs pollinated by bumblebees were also examined. Our preliminary results indeed

suggest that the fruit production of understorey herbs is affected by the flowering intensity of canopy

trees via the activities of pollinators (Fig. 54; contact: Tsutom Hiura).

98 99

Background and management system

The Wind River Canopy Crane was erected in April of 1995 by the

University of Washington, in cooperation with the United States

Department of Agriculture, Forest Service, Pacific Northwest Research

Station (Portland, Oregon), Wind River Experimental Forest (Skamania

County, Washington) and Gifford Pinchot National Forest (Vancouver,

Washington). The crane is owned and operated by the University of

Washington. The Director is Prof. Jerry Franklin (Univ. Wash.). Dr.

Rick Meinzer (USFS) is Co-PI and Canopy Team Leader at the Pacific

Northwest Research Station. Funding for the project comes from the

USDA Forest Service, Pacific Northwest Research Station and the

University of Washington.

The crane is 74.5m tall at the jib, with an arm of 85m (Figs 55 and 56).

The surrounding forest attains a maximum height of 63m, and the crane

can gain access to 320 canopy trees in the 2.3ha perimeter. There are

seven conifer and two angiosperm tree species in the crane perimeter.

The crane was placed off an abandoned haul road through the old-growth

forest using a mobile crane. Power is provided from access to line systems

approximately 1.5km distant.

The crane research facility is managed by the Director with support from

a two tiered committee system; an Operations Committee which meets

bi-monthly to direct day to day management of the research facility and

evaluate proposed research, and a National Scientific Advisory

Committee which meets once a year to provide scientific review of the

overall management and mission. Facilities at this field station include

office, shop, and storage buildings, and two bunk houses that can

accommodate 12 and 8 people. There are five full time staff on site: Research Manager (D.C. Shaw),

Research Coordinator, Program Coordinator, Research Tower Crane Operator, and Field Meteorological

Scientist. The Director is based in Seattle at the University of Washington. Co-PI is located in Corvallis,

Oregon at the Pacific Northwest Research Station, Forestry Sciences Laboratory.

The Wind River Canopy Crane Research Facility (WRCCRF) is managed as a user facility. Researchers

and educators make proposals via our web site (http://depts.washington.edu/wrccrf). Research proposals

are reviewed by a sub-group of the Operations Committee. Education proposals, usually requesting an

‘educational lift’ (i.e., an educational tour into the canopy), are handled by the Research Manager.

Educational lifts are provided for college classes, academic and professional societies, agency personnel,

and the general public through adult education classes. Our web page provides researchers, academics,

and others access to information regarding the research facility, our program and some research findings.

The WRCCRF is not generally accessible to the public, although an interpretative trail does go near the

border of our research area, and there is a place to view the crane through the canopy, along with an

interpretative sign.

Description of the site

The canopy crane is located in the Thornton T. Munger Research Natural Area (Table 12, Figs 57 and

58), a 478ha old-growth (500yr) forest. This low elevation (335-610m) Pseudotsuga menziezii (Douglas-

fir) / Tsuga heterophylla (western hemlock) forest was designated for non-destructive research by the US

Forest Service in 1934 to represent the typical forest type that originally covered many valleys in western

Washington’s Cascade Mountain Range. The Research Natural Area is located in the Wind River

Experimental Forest, managed by the US Forest Service Pacific Northwest Research Station. The

Experimental forest (4,208ha) has a range of forest ages and cutting regimes and is the site of

chronosequence studies linking the canopy crane old-growth to younger forest types.

4.2.6. Wind River Canopy Crane Research Facility, USADavid C. Shaw, Frederick C. Meinzer, Ken J. Bible & Geoffrey G. Parker

Fig. 55. Wind River Canopy Crane (photo Jerry Franklin).

Fig. 56. Wind River Canopy Crane (photo Jerry Franklin).

Variable Characteristics

Location T. T. Munger Research Natural Area, Gifford Pinchot National Forest, WashingtonState, USA, 45°49' 13.76" N, 121°57' 06.88" W

Altitude 335mMean annual air temperature 8.7˚CMean annual rainfall 2197mmType of forest Temperate coniferous seasonal rainforestArea of forest accessed by the crane 2.3haCanopy height Max. 63mCrane model Liebherr 550EC, fixedHeight of tower / Length of jib 74.5m / 85mMaximum height reached by the gondola ca. 67mGondola type a: Squared, 1.2 x 1.2 x 2.2m, max. load 454kg

b: Rectangular, 2.7 x 1.3 x 2.2m, max. load 908kgNumber of persons carried by the gondola a: 4 persons

b: 8 personsIn operation since 1995Main research topics • Forest structure and diversity

• Biodiversity• Parasitic plants• Tree physiology• Climate change and carbon dynamics

Remarks Educational crane lifts may be provided for academic and professional societies,and for the public through adult education classes

Management University of WashingtonContact David C. Shaw, University of Washington

E-mail dshaw@u.washington.eduWeb site http://depts.washington.edu/wrccrf/List of publications http://depts.washington.edu/wrccrf/Fees for researchers USD 182 per hour, negotiable

Table 12. Site and crane characteristics of the Wind River Canopy Crane Research Facility.

Climate

The canopy crane is located in a temperate coniferous seasonal rainforest (Shaw et al., in press) with four

distinct seasons (Table 13, monthly averages). Average annual precipitation is 2,197mm, with a distinct

dry period occurring during June, July and August. This ‘summer drought’ is an important characteristic

of Pacific Northwest climate, since it affects summertime tree growth and productivity. Average annual

temperature is 8.7ºC. Much of the winter precipitation falls as snow (average snowfall 2,330mm), and we

experience rain-on-snow events commonly during the winter months.

100 101

The WRCCRF is located on the crest of the Cascade Mountain Range

where continental climates dominate to the east, while maritime climate

dominates to the west. Winter weather is typically maritime, whilst

summer weather is typically continental. Fifteen kilometers to the south

is the Columbia River Gorge, which has dramatic weather patterns caused

by dynamics of deserts in the east and ocean to the west. In winter,

freezing rains, rain-on-snow, and hail events are common.

Forest history

It is probable that the forest stand in the T.T. Munger RNA originated

after a fire or series of fires about 500 years ago. Pseudotsuga menziesii

(Fig. 59) dominated the young forest for the first 200 years, after which,

the shade tolerant Tsuga heterophylla (Fig. 60) and Thuja plicata (western

redcedar) began to move into the understorey. Pseudotsuga menziesii still

maintains dominance in basal area, wood volume and height. The forest

is unusually diverse for the Cascade Mountains with eight coniferous

species, including P. menziesii (Pinaceae), T. heterophylla (Pinaceae), T.

plicata (Cupressaceae), Taxus brevifolia (Pacific yew, Taxaceae), Abies

amabilis (Pacific silver fir, Pinaceae), A. procera (noble fir, Pinaceae), A.

grandis (grand fir, Pinaceae), and Pinus monticola (western white pine,

Pinaceae).

Flora and fauna

Vascular and non-vascular plant, lichen and macrofungi species diversity

surveys of the 4 ha site under the crane have been completed or are in

progress. To date, we have identified 75 vascular plant species, including

10 trees (8 conifers, 2 angiosperms), 14 shrubs, 40 herbs, 5 graminoids,

and 6 ferns. In addition there are roughly 52 mosses, 7 liverworts and

113 (+) lichens. The epiphytic plant community is exclusively non-

vascular plants, bryophytes and lichens. There is a strong stratification

of the epiphytic community with bryophytes dominating the lower

canopy, cyanolichens (nitrogen fixers) dominating the mid canopy, and

green-algal-foliose and alectorioid lichens dominating the upper canopy.

Highest bryophyte diversity is on the forest floor and on fallen timber.

Interestingly, the only angiosperm above 15m is Arceuthobium tsugense

(hemlock dwarf mistletoe, Viscaceae), a parasite on T. heterophylla that

is common over about 1ha of the 4ha plot. The genus Arceuthobium

differs from other mistletoes in that it spreads by explosively discharged

seed, rather than bird dispersal, which makes the composition and

structure of the forest a controlling influence on its spread.

Thirty-three plant families are represented in the vascular plants, 18 of

which have only one species, and 7 families have 2 species. Dominant

plant families include Ericaceae (8 spp.), Liliaceae (7 spp.), Pinaceace

(6 spp.), Polypodiaceae (5 spp.) and Roseaceae (5 spp.). Understorey

vegetation cover and frequency were quantified on sixty-four 25x25 m

plots in the 4 ha study area. Prominence values (pv) were determined by

multiplying the average percent cover by the % frequency. The dominant

understorey species were Acer circinatum (Aceraceae, pv 25.5, Fig. 61),

Gaultheria shallon (Ericaceae, pv 16.0), Berberis nervosa (Berberidaceae,

pv 14.2), Achlys triphylla (Berberidaceae, pv 4.4), Vaccinium parvifolium

(Ericaceae, pv 4.2), Vancouvaria hexandra (Berberidaceae, pv 1.4)

Pteridium aquilinum (Polypodiaceae, pv 1.4), Linnaea borealis

(Caprifoliaceae, pv 1.1), and Xerophyllum tenax (Liliaceae, pv 1.0). No

other plant species had a pv of 1 or greater.

Fifty species of birds have been identified in on-going avian research.

Seventeen of these are Neotropical migrants, while the remainder are

residents or regional migrants. The most common residents are

Troglodytes troglodytes (Winter Wren), Poecile rufescens (Chestnut-

backed Chickadee), Certhia americana (Brown Creeper), Perisoreus

canadensis (Gray Jay), Turdus migratorius (American Robin), and Regulus

satrapa (Golden-crowned Kinglet). The most abundant neotropical

migrants are Empidonax difficilis (Pacific-slope Flycatcher), Catharus

guttatus (Hermit Thrush), Chordeiles minor (Common Nighthawk),

Catharus ustulatus (Swainson’s Thrush), and Dendroica occidentalis

(Hermit warbler). Larger birds using the site include Dryocopus pileatus

(Pileated Woodpecker), Corvus corax (Raven), Strix varia (Barred Owl),

and Accipiter gentilis (Goshawk). Strix occidentalis (Spotted Owl) has

apparently been displaced by S. varia. Sturnus vulgaris (European

Starling) have been nesting on the crane, which we attempt to control.

Reptiles are rare in this forest type, and include two snakes Thamnophis

sirtalis (Common Garter Snake) and Charina bottae (Rubber Boa).

Amphibians include Ensatina escholtzii (Ensatina), Ambystoma

macrodactylum (Long-toed Salamander), Ambystoma gracile

(Northwestern Salamander), Taricha granulosa (Roughskin Newt), Rana

aurora (Red-legged Frog), and Hyla regilla (Pacific Tree Frog).

Thirty-nine species of mammals are thought to occur in the area around

the Research Natural Area, in 6 orders, including Insectivora (5 spp.),

Chiroptera (9 spp.), Lagomorpha (1 spp.), Rodentia (13 spp.), Carnivora

(9 spp.), and Artiodactyla (2 spp.). Most notable of these is Cervus

canadensis (elk), Felis concolor (cougar), and Ursus americanus (black

bear). There are no exclusively arboreal mammals in this forest, the most

‘arboreal’ of the mammals are Glaucomys sabrinus (northern flying

squirrel), Tamiasciurus douglasii (Douglas’ squirrel), Neotoma cinerea

(bushy-tailed woodrat), Tamias townsendii (Townsend’s chipmunk),

Erethizon dorsatum (porcupine), Martes americaana (martin), and Martes

pennanti (fisher).

Fig. 57. Wind River Canopy Crane looking west over T.T. Munger ResearchNatural Area. The crane is in the lower right corner (photo David Shaw).

Fig. 58. Wind River Valley looking south. The crane is in the centre of thephoto, just this side of the brown agricultural fields. Mt. Hood is in the backleft (photo David Shaw).

Table 13. Long-term microclimate summary for the WRCCRF derived fromdata collected between 1978 and 2001, inclusive, at the Carson Fish Hatchery,a National Oceanic and Atmospheric Administration (www.ncdc.noaa.gov)meteorological station located 5.7 km NNW from the canopy crane.

Fig. 59. Pseudotsuga menziesii new foliage flush from above (photo JerryFranklin).

Fig. 60. Snow on Tsuga heterophylla (photo Andrew Baker).

Fig. 61. Acer circinatum in the understorey(photo Jerry Franklin).

102 103

Stand characteristics

The forest stand has been characterized on a 4ha plot, which has been

expanded to a 12ha plot. However, data presented here are based on the

4ha plot directly under and around the canopy crane, with all trees with

diameter at breast height (dbh) ≥ 5cm mapped. The basal area is 82.8

m2/ha, while trees average 428 stems/ha. For trees with dbh ≥ 10cm,

density is 309 stems/ha, while for trees with dbh ≥ 20cm, density is 194

stems/ha. It is a classic late-successional forest with a population of large

and relatively evenly aged P. menziesii dominating the forest in terms of

height, wood volume and basal area. These trees are dying out of the

stand slowly and there is no reproduction of these shade-intolerant trees

in the understorey. T. heterophylla, T. plicata, and A. amabilis are shade

tolerant, dominate the smaller diameter classes, and represent the primary

reproducing species (Fig. 62). Taxus brevifolia is an important understorey, small stature tree, averaging

89 trees/ha. The density of trees in 5cm diameter classes is summarized in Table 14, and reflects the

abundance of small diameter trees in this uneven aged old-growth forest. Height distributions are

summarized in Figure 63. Average height of P. menziesii is 52.2m (tallest tree is 65m), whilst T. heterophylla

averages 19.0m (tallest tree is 55m). Pseudotsuga menziesii has only 8.5% of the stems present but 43.3%

of the basal area, 47.9% of the stem wood volume, and 33% of foliage. However, Tsuga heterophylla has

51.7% of the stems, with 31.4% of the basal area, 32.4% of the stem wood volume, and 45.7% of the

foliage. Thuja plicata has 6.9% of the stems, 19.9% of the basal area, 17% of the stem wood volume, and

15.4% of the foliage.

Main research topics studied at the site

a) Forest carbon, water and nutrient cycles (biogeochemical cycles)

Developing science-based strategies for managing biogeochemical cycles will require much additional

work to increase our understanding of them, the controls on them, interactions among them, and how

they are being perturbed by human activities. Quantifying sources and sinks for carbon is particularly

important from the standpoint of predicting and ameliorating carbon dioxide-driven global climate change.

Since the carbon cycle is linked inextricably with the water and nutrient cycles, these cycles must be

studied concurrently in order to arrive at a full understanding of factors controlling the magnitude of

carbon dioxide sources and sinks.

Relevant research at the WRCCRF includes a major interdisciplinary initiative supported by the Western

Region of the National Institute for Global Environmental Change. Multiple teams of investigators are

employing independent approaches to determine the magnitude of carbon and water fluxes and to

characterize the mechanisms that control them. These activities have resulted in Wind River being

designated as an official Ameriflux site (part of the FLUXNET network, see Chapter 2). In the Ameriflux

network, a micrometeorological technique known as eddy covariance is being used to measure canopy

level fluxes of carbon dioxide and water vapour continuously in a range of vegetation types. These

measurements will help identify major carbon sources and sinks.

Whole stand estimates of tree water use are being attempted using sap

flow sensors. This will provide data on quantities of material transfer

with the atmosphere. Nitrogen cycling is being studied below ground

and at multiple canopy levels.

b) Biological diversity and ecosystem functioning

The maintenance of biological diversity is a key goal of the US Forest

Service. Biological diversity is known to affect ecosystem functioning,

with rare and endangered organisms being considered potential indicators

of problems. The spatial and temporal organization of biological

organisms within a forest community is a key factor in understanding

how its biological diversity may be sustained.

The WRCCRF is a major new tool being used to investigate how biological

organisms are organized in a tall stature, old-growth conifer forest. Organisms being studied include

insects, fungi, birds, bats, small mammals, lichens, and mosses. Ecosystem processes dependent on

biological diversity that are being studied concurrently include nitrogen fixation, forest hydrology, avian

controls on leaf herbivory, dwarf mistletoe disease development, and food resources for vertebrates.

c) Forest health and protection (plant pathology and disease)

Forest health is a major societal concern, and keeping forests healthy is a goal of the US Forest Service in

its management of forest lands. For example, the Forest Service has determined that dwarf mistletoe has

a greater impact on productivity in western forests than insects and diseases combined.

The WRCCRF has provided a unique opportunity for scientists to study crown-dwelling pest organisms

that would be difficult to survey. Since old-growth forests in western Washington and Oregon are known

for their lack of defoliating insect outbreaks, researchers at the canopy crane are studying what controls

herbivory by insects in the Wind River old-growth forest. A major research initiative on dwarf mistletoe

epidemiology has also been conducted at the canopy crane since it was installed. Finally, the WRCCRF

has partnered with the Swiss Needle Cast Cooperative, and the canopy crane is now used as a reference

site to study and monitor a natural population of Swiss Needle Cast (Phaeocryptopus gaumannii), a fungi

causing early needle loss and reduced growth of Douglas-fir in the coast range of Oregon and Washington.

d) Monitoring of climate and climate variability

Assessing and predicting the impact of climate variability on forests requires continuous monitoring of

both basic climate variables and forest response variables. Basic climatic monitoring at the WRCCRF is

accomplished with an open field meteorological station and a vertical array of sensors on the crane tower

for characterizing the microclimate at different depths in the forest canopy. Forest response variables

being measured include phenology, diameter growth, annual mortality, cone crops, and needle longevity.

Fig. 62. Number of trees by diameter class for the 4ha plot around the WindRiver Canopy Crane.

dbh

From To Trees/ha

5.0 9.9 119.3

10.0 14.9 74.3

15.0 19.9 41.3

20.0 24.9 28.0

25.0 29.9 16.8

30.0 34.9 16.0

35.0 39.9 13.3

40.0 44.9 8.8

45.0 49.9 7.0

50.0 54.9 7.5

55.0 59.9 8.5

60.0 64.9 7.8

65.0 69.9 7.0

70.0 74.9 7.3

75.0 79.9 7.0

80.0 84.9 7.3

85.0 89.9 6.8

90.0 94.9 4.3

95.0 99.9 4.8

100.0 104.9 4.8

105.0 109.9 5.8

110.0 114.9 4.5

115.0 119.9 3.3

120.0 124.9 3.3

125.0 129.9 4.5

130.0 134.9 1.5

135.0 139.9 1.3

140.0 144.9 2.3

145.0 149.9 1.5

150.0 154.9 0.8

155.0 159.9 0.8

Table 14. Total numberof trees (all species) in 5cmdiameter classes.

Fig. 63. Number of trees by height class for the 4ha plot around the WindRiver Canopy Crane.

104 105

e) Validation and testing of new remote sensing technology

The development of remote sensing technology has great potential for facilitating rapid and frequent

assessment of the status of forest stands with regard to important variables such as health and productivity.

However, ground-truthing of remote sensing interpretation has been hampered by difficulty in accessing

the forest canopy, which contributes most of the reflected light detected by remote sensing devices.

Canopy access tools such as cranes are playing an important role in overcoming this impediment.

At the WRCCRF a multi-institutional team has been exploring ways of linking forest properties with

remotely sensed spectral data. This ground-truthing effort will be instrumental in attaining the ultimate

goal of using remote sensing data to provide stand level, regional, and global assessments of the status of

forests and other vegetation.

f) Tree physiology and growth

Due to their large size and logistical difficulties in gaining access to their crowns, physiological processes

in trees have traditionally been studied at a single scale or over a limited range of scale. The availability

of large canopy cranes, has presented new opportunities for studying large trees as whole, integrated

organisms. An understanding of how processes such as photosynthesis (Fig. 64) and water use are

integrated at the whole-tree level is essential for linking different scales in ecological process models

effectively and for a seamless understanding of how processes are controlled as we move along a continuum

of scale from leaves to ecosystems.

Research on tree physiology at the WRCCRF has concentrated on control of water uptake, transport and

loss along the soil-to-atmosphere continuum, seasonal, spatial and leaf age-related patterns of canopy

photosynthesis, and relationships between tree architecture and growth. Future work will concentrate

on tree physiological behaviour along a chronosequence of forest age in the Wind River Experimental

Forest.

Major findings

a) Forest carbon, water and nutrient cycles (biogeochemical cycles)

The total store of carbon in living plants at the Wind River site was 39,800g C m-2 (Harmon et al., in

press). Long term, the site is considered a slight sink for carbon dioxide, yet contrary to conventional

wisdom, continuous measurements of canopy carbon dioxide flux at the WRCCRF have shown that old-

growth coniferous forests can be significant carbon dioxide sinks in certain years (Paw U et al., in press).

The fact that old-growth forests can function as carbon storehouses rather than carbon sources has far-

reaching implications for managing forests to improve carbon sequestration. Future work at WRCCRF

and surrounding Wind River Experimental Forest will concentrate on carbon sequestration as a function

of forest age.

A special issue of the journal ‘Ecosystems’ will focus on the carbon program and these papers are in

press (Suchanek et al., in press, lead paper). A special issue of the journal ‘Tree Physiology’ (Vol. 22 No.

2 and 3) included many papers that have focused on connecting the canopy crane site to younger forests

in a chronosequence (Bond & Franklin, 2002, lead paper).

b) Biological diversity and ecosystem functioning

Work on biodiversity and ecosystem functioning has identified specific elements of forest structure that

contribute to biodiversity. For many types of organisms, biodiversity in the old-growth forest at the

WRCCRF has a strong vertical component that outweighs other structural features of the forest. For

example, lichen diversity increases but moss diversity decreases with increasing height in the canopy. A

unique, previously undescribed lichen community associated with dead tops and roosting posts of birds

has been discovered in the upper 2m of the canopy (McCune et al., 2000). Conversely, arthropod diversity

is determined largely by the species of the host tree. A unique arthropod community dominated by mites

and spiders is associated with western red cedar (Schowalter & Ganio, 1998). This result implies that

forests managed for fewer tree species eliminate important components of arthropod diversity. Vertically

stratified avian surveys have shown that during March - October there are equal numbers of birds detected

in the lower, mid, and upper canopy, but during November - February (winter) there is a distinct shift in

detections to the upper canopy. This implies that the structure and resources of the upper canopy of old-

growth forests are more important for birds than previously realized (Shaw et al., 2002).

c) Forest health and protection (plant pathology and disease)

Research on hemlock dwarf mistletoe (Arceuthobium tsugense, Viscaceae) epidemiology is ongoing at

WRCCF and has shown that vertical structure of the canopy and species composition of the forest are the

major factors controlling the intensification and spread of infection. Plants tend to be clumped into

infection centres that are expanding as western hemlock increases in abundance and importance. Herbivory

research has shown that in the overstorey conifers leaf area loss to insect herbivores is less that 2%, while

in understorey vine maple (Acer circinatum) it was 10% (Braun et al., 2002). Research on Swiss needle

cast at WRCCF is employing an assay that allows total fungal biomass to be estimated based on extraction

of mitochondrial DNA from infected needles. This assay will facilitate rapid assessment of the severity of

Swiss needle cast infection, whilst current research is attempting to quantify the impacts of this disease

on carbon dynamics of Douglas-fir.

d) Monitoring of climate and climate variability

The continuous records of vertical microclimate being obtained probably represent the only such data

available for old-growth coniferous forests. These data are valuable to researchers working on biological

diversity, forest health, and other topics, and will help us understand the influence of forests on local and

regional climate. The data are available to all researchers via the Internet (http://depts.washington.edu/

wrccrf).

e) Ground validation and testing of new remote sensing technology

The vertical structure of the forest canopy, as well as all the traditional aspects of forest structure (dbh

distribution, basal area, heights) have been quantified in a core 4ha plot around the canopy crane. This

uniquely detailed information on forest structure has allowed the canopy crane forest to serve as a major

site for validation of data quality (Freeman & Ford, 2002), forest mistletoe surveys (Shaw et al., 2000)

and new remote sensing technologies being developed and field-tested in the United States today (Parker

et al., 2001; Lefsky et al., 2002). Ultra-lights, airplanes, and satellites have overflown the site, and even

the canopy crane structure itself has had remote sensing instrumentation attached to it 15m above the

Fig. 64. William Winner and SeanThomas measuring photosynthesis fromgondola (photo Andrew Baker).

106 107

forest canopy. These validations either would not have occurred at the WRCCRF, or in many cases,

would not have been possible at all without the availability of the crane.

f) Tree physiology and growth

Research on water uptake and utilization by trees has shown that Douglas-fir roots are able to redistribute

hydraulically water from wetter to drier portions of the soil profile (Brooks et al., 2002) and that large

trees rely on rechargeable stem water storage for a larger fraction of their daily water use than small trees

(McDowell et al., 2002). Both of these findings have important implications for adaptation of Douglas-

fir to summer drought, for ecosystem-level hydrology, and, therefore, yield of watersheds occupied by

stands of different ages. Needle-level light response of photosynthesis across a vertical light gradient has

shown that mean light-saturated photosynthetic rates, light compensation points, and respiration rates

declined from upper crowns to lower crowns in overstorey Douglas-fir and western hemlock. Increasing

light-saturated photosynthetic rates in the canopy top increased carbon uptake at high photosynthetic

photon flux density (PPFD) relative to lower canopy needles. Reduced respiration rates in lower canopy

needles relative to upper canopy needles increased carbon uptake at low PPFD by reducing the light

compensation point (Lewis et al., 2000).

With regard to growth, about 20 to 70 per cent of the foliage on old-growth Douglas-fir trees is found on

epicormic branches. These are non-primary branches that sprout from dormant buds in the trunk and on

other branches. This branching behaviour is a major factor contributing to the longevity of Douglas-fir

(Ishii & Ford, 2001). The leaf area index (LAI) of this forest is estimated at 8.61, while 20% (1.69) is in

the understorey vegetation (Thomas & Winner, 2000). The canopy crane was used in a novel fashion to

estimate this LAI.

g) Forest structure

The development of forest structure and its influence on microclimate

have a controlling influence on the productivity and biodiversity of forest

ecosystems. The Wind River old-growth forest is ‘bottom heavy’ with

structure, meaning the largest amount of material is located below 30m

(Parker, 1997). The light environment has been divided into three zones,

a bright zone, where high light dominates, above 40m, a transition zone,

where the mean light transmittance changes rapidly with height (40-12m),

and a dim zone which is characterized by low transmittance (< 12m; Fig.

65). When compared to stands in a chronosequence, the bright zone is

wider and dim zone is narrower in old-growth than younger forests

(Parker et al., 2002). The vertical stratification of species in the Wind

River old-growth forest allows Douglas-fir to maintain dominance and

longevity (Ishii et al., 2000), whilst tree age is significantly related to the

development of crown structural parameters (Ishii & McDowell, 2002).

Future visions and collaboration with the ICCN

The Wind River Canopy Crane will continue to be an ecosystem observatory, a research tool for a broad

array of scientists, and a facilitator for education. We hope the research program broadens to include

forests of all ages (chronosequence) and structures (land management regimes) in an attempt to understand

forest ecosystem processes at all scales and provide context for the Wind River old-growth forest and

canopy crane research program. Climate change and carbon dynamics will be a major focus of long term

research, which will integrate nutrient and water cycles. We hope to integrate other research more closely

with National and International environmental observation networks, which are critical to understanding

global change. Whole forest processes, such as controls on net primary productivity, carbon cycling,

photosynthesis, and water use will continue to be central to our research program. Finally, understanding

the foundations of biodiversity in forests, particularly the role of the hard-to-access upper canopy, will

continue as a central topic of research.

Opportunities to collaborate with the International Canopy Crane Network are numerous, and at various

scales. Collaboration may be as simple as hosting visiting researchers who are also doing research at

other canopy cranes, or as complex as joint research grants or programs. Several areas stand out, including

linking the canopy cranes into an observation network, addressing key scientific questions about

biodiversity and ecosystem process, testing hypotheses at multiple crane sites, and elucidating ‘universal

rules’ that apply to tree physiology, canopy processes and forest health across multiple forest biomes.

Fig. 65. Variation in PAR transmittance profiles with height in the old-growthforest under the Wind River Canopy Crane.